Datasets:

CShorten

/

Last-Week-on-ML-ArXiv

like 0

TNN7: A Custom Macro Suite for Implementing Highly Optimized Designs of Neuromorphic TNNs

Temporal Neural Networks (TNNs), inspired from the mammalian neocortex, exhibit energy-efficient online sensory processing capabilities. Recent works have proposed a microarchitecture framework for implementing TNNs and demonstrated competitive performance on vision and time-series applications. Building on these previous works, this work proposes TNN7, a suite of nine highly optimized custom macros developed using a predictive 7nm Process Design Kit (PDK), to enhance the efficiency, modularity and flexibility of the TNN design framework. TNN prototypes for two applications are used for evaluation of TNN7. An unsupervised time-series clustering TNN delivering competitive performance can be implemented within 40 uW power and 0.05 mm^2 area, while a 4-layer TNN that achieves an MNIST error rate of 1% consumes only 18 mW and 24.63 mm^2. On average, the proposed macros reduce power, delay, area, and energy-delay product by 14%, 16%, 28%, and 45%, respectively. Furthermore, employing TNN7 significantly reduces the synthesis runtime of TNN designs (by more than 3x), allowing for highly-scaled TNN implementations to be realized.

Training neural networks using Metropolis Monte Carlo and an adaptive variant

We examine the zero-temperature Metropolis Monte Carlo algorithm as a tool for training a neural network by minimizing a loss function. We find that, as expected on theoretical grounds and shown empirically by other authors, Metropolis Monte Carlo can train a neural net with an accuracy comparable to that of gradient descent, if not necessarily as quickly. The Metropolis algorithm does not fail automatically when the number of parameters of a neural network is large. It can fail when a neural network's structure or neuron activations are strongly heterogenous, and we introduce an adaptive Monte Carlo algorithm, aMC, to overcome these limitations. The intrinsic stochasticity of the Monte Carlo method allows aMC to train neural networks in which the gradient is too small to allow training by gradient descent. We suggest that, as for molecular simulation, Monte Carlo methods offer a complement to gradient-based methods for training neural networks, allowing access to a distinct set of network architectures and principles.

What GPT Knows About Who is Who

Coreference resolution -- which is a crucial task for understanding discourse and language at large -- has yet to witness widespread benefits from large language models (LLMs). Moreover, coreference resolution systems largely rely on supervised labels, which are highly expensive and difficult to annotate, thus making it ripe for prompt engineering. In this paper, we introduce a QA-based prompt-engineering method and discern \textit{generative}, pre-trained LLMs' abilities and limitations toward the task of coreference resolution. Our experiments show that GPT-2 and GPT-Neo can return valid answers, but that their capabilities to identify coreferent mentions are limited and prompt-sensitive, leading to inconsistent results.

SuperWarp: Supervised Learning and Warping on U-Net for Invariant Subvoxel-Precise Registration

In recent years, learning-based image registration methods have gradually moved away from direct supervision with target warps to instead use self-supervision, with excellent results in several registration benchmarks. These approaches utilize a loss function that penalizes the intensity differences between the fixed and moving images, along with a suitable regularizer on the deformation. In this paper, we argue that the relative failure of supervised registration approaches can in part be blamed on the use of regular U-Nets, which are jointly tasked with feature extraction, feature matching, and estimation of deformation. We introduce one simple but crucial modification to the U-Net that disentangles feature extraction and matching from deformation prediction, allowing the U-Net to warp the features, across levels, as the deformation field is evolved. With this modification, direct supervision using target warps begins to outperform self-supervision approaches that require segmentations, presenting new directions for registration when images do not have segmentations. We hope that our findings in this preliminary workshop paper will re-ignite research interest in supervised image registration techniques. Our code is publicly available from https://github.com/balbasty/superwarp.

Sibyl: Adaptive and Extensible Data Placement in Hybrid Storage Systems Using Online Reinforcement Learning

Hybrid storage systems (HSS) use multiple different storage devices to provide high and scalable storage capacity at high performance. Recent research proposes various techniques that aim to accurately identify performance-critical data to place it in a "best-fit" storage device. Unfortunately, most of these techniques are rigid, which (1) limits their adaptivity to perform well for a wide range of workloads and storage device configurations, and (2) makes it difficult for designers to extend these techniques to different storage system configurations (e.g., with a different number or different types of storage devices) than the configuration they are designed for. We introduce Sibyl, the first technique that uses reinforcement learning for data placement in hybrid storage systems. Sibyl observes different features of the running workload as well as the storage devices to make system-aware data placement decisions. For every decision it makes, Sibyl receives a reward from the system that it uses to evaluate the long-term performance impact of its decision and continuously optimizes its data placement policy online. We implement Sibyl on real systems with various HSS configurations. Our results show that Sibyl provides 21.6%/19.9% performance improvement in a performance-oriented/cost-oriented HSS configuration compared to the best previous data placement technique. Our evaluation using an HSS configuration with three different storage devices shows that Sibyl outperforms the state-of-the-art data placement policy by 23.9%-48.2%, while significantly reducing the system architect's burden in designing a data placement mechanism that can simultaneously incorporate three storage devices. We show that Sibyl achieves 80% of the performance of an oracle policy that has complete knowledge of future access patterns while incurring a very modest storage overhead of only 124.4 KiB.

Inverse design of nano-photonic wavelength demultiplexer with a deep neural network approach

In this paper, we propose a pre-trained-combined neural network (PTCN) as a comprehensive solution to the inverse design of an integrated photonic circuit. By utilizing both the initially pre-trained inverse and forward model with a joint training process, our PTCN model shows remarkable tolerance to the quantity and quality of the training data. As a proof of concept demonstration, the inverse design of a wavelength demultiplexer is used to verify the effectiveness of the PTCN model. The correlation coefficient of the prediction by the presented PTCN model remains greater than 0.974 even when the size of training data is decreased to 17%. The experimental results show a good agreement with predictions, and demonstrate a wavelength demultiplexer with an ultra-compact footprint, a high transmission efficiency with a transmission loss of -2dB, a low reflection of -10dB, and low crosstalk around -7dB simultaneously.

Learning Representations for New Sound Classes With Continual Self-Supervised Learning

In this paper, we present a self-supervised learning framework for continually learning representations for new sound classes. The proposed system relies on a continually trained neural encoder that is trained with similarity-based learning objectives without using labels. We show that representations learned with the proposed method generalize better and are less susceptible to catastrophic forgetting than fully-supervised approaches. Remarkably, our technique does not store past data or models and is more computationally efficient than distillation-based methods. To accurately assess the system performance, in addition to using existing protocols, we propose two realistic evaluation protocols that use only a small amount of labeled data to simulate practical use cases.

Incorporating Prior Knowledge into Neural Networks through an Implicit Composite Kernel

It is challenging to guide neural network (NN) learning with prior knowledge. In contrast, many known properties, such as spatial smoothness or seasonality, are straightforward to model by choosing an appropriate kernel in a Gaussian process (GP). Many deep learning applications could be enhanced by modeling such known properties. For example, convolutional neural networks (CNNs) are frequently used in remote sensing, which is subject to strong seasonal effects. We propose to blend the strengths of deep learning and the clear modeling capabilities of GPs by using a composite kernel that combines a kernel implicitly defined by a neural network with a second kernel function chosen to model known properties (e.g., seasonality). Then, we approximate the resultant GP by combining a deep network and an efficient mapping based on the Nystrom approximation, which we call Implicit Composite Kernel (ICK). ICK is flexible and can be used to include prior information in neural networks in many applications. We demonstrate the strength of our framework by showing its superior performance and flexibility on both synthetic and real-world data sets. The code is available at: https://anonymous.4open.science/r/ICK_NNGP-17C5/.

The Splendors and Miseries of Heavisidisation

Machine Learning (ML) is applicable to scientific problems, i.e. to those which have a well defined answer, only if this answer can be brought to a peculiar form ${\cal G}: X\longrightarrow Z$ with ${\cal G}(\vec x)$ expressed as a combination of iterated Heaviside functions. At present it is far from obvious, if and when such representations exist, what are the obstacles and, if they are absent, what are the ways to convert the known formulas into this form. This gives rise to a program of reformulation of ordinary science in such terms -- which sounds like a strong enhancement of the constructive mathematics approach, only this time it concerns all natural sciences. We describe the first steps on this long way.

Effect of Batch Normalization on Noise Resistant Property of Deep Learning Models

The fast execution speed and energy efficiency of analog hardware has made them a strong contender for deployment of deep learning model at the edge. However, there are concerns about the presence of analog noise which causes changes to the weight of the models, leading to performance degradation of deep learning model, despite their inherent noise resistant characteristics. The effect of the popular batch normalization layer on the noise resistant ability of deep learning model is investigated in this work. This systematic study has been carried out by first training different models with and without batch normalization layer on CIFAR10 and CIFAR100 dataset. The weights of the resulting models are then injected with analog noise and the performance of the models on the test dataset is obtained and compared. The results show that the presence of batch normalization layer negatively impacts noise resistant property of deep learning model and the impact grows with the increase of the number of batch normalization layers.

High-Resolution CMB Lensing Reconstruction with Deep Learning

Next-generation cosmic microwave background (CMB) surveys are expected to provide valuable information about the primordial universe by creating maps of the mass along the line of sight. Traditional tools for creating these lensing convergence maps include the quadratic estimator and the maximum likelihood based iterative estimator. Here, we apply a generative adversarial network (GAN) to reconstruct the lensing convergence field. We compare our results with a previous deep learning approach -- Residual-UNet -- and discuss the pros and cons of each. In the process, we use training sets generated by a variety of power spectra, rather than the one used in testing the methods.

What is an equivariant neural network?

We explain equivariant neural networks, a notion underlying breakthroughs in machine learning from deep convolutional neural networks for computer vision to AlphaFold 2 for protein structure prediction, without assuming knowledge of equivariance or neural networks. The basic mathematical ideas are simple but are often obscured by engineering complications that come with practical realizations. We extract and focus on the mathematical aspects, and limit ourselves to a cursory treatment of the engineering issues at the end.

Novel Multicolumn Kernel Extreme Learning Machine for Food Detection via Optimal Features from CNN

Automatic food detection is an emerging topic of interest due to its wide array of applications ranging from detecting food images on social media platforms to filtering non-food photos from the users in dietary assessment apps. Recently, during the COVID-19 pandemic, it has facilitated enforcing an eating ban by automatically detecting eating activities from cameras in public places. Therefore, to tackle the challenge of recognizing food images with high accuracy, we proposed the idea of a hybrid framework for extracting and selecting optimal features from an efficient neural network. There on, a nonlinear classifier is employed to discriminate between linearly inseparable feature vectors with great precision. In line with this idea, our method extracts features from MobileNetV3, selects an optimal subset of attributes by using Shapley Additive exPlanations (SHAP) values, and exploits kernel extreme learning machine (KELM) due to its nonlinear decision boundary and good generalization ability. However, KELM suffers from the 'curse of dimensionality problem' for large datasets due to the complex computation of kernel matrix with large numbers of hidden nodes. We solved this problem by proposing a novel multicolumn kernel extreme learning machine (MCKELM) which exploited the k-d tree algorithm to divide data into N subsets and trains separate KELM on each subset of data. Then, the method incorporates KELM classifiers into parallel structures and selects the top k nearest subsets during testing by using the k-d tree search for classifying input instead of the whole network. For evaluating a proposed framework large food/non-food dataset is prepared using nine publically available datasets. Experimental results showed the superiority of our method on an integrated set of measures while solving the problem of 'curse of dimensionality in KELM for large datasets.

Learning Car Speed Using Inertial Sensors

A deep neural network (DNN) is trained to estimate the speed of a car driving in an urban area using as input a stream of measurements from a low-cost six-axis inertial measurement unit (IMU). Three hours of data was collected by driving through the city of Ashdod, Israel in a car equipped with a global navigation satellite system (GNSS) real time kinematic (RTK) positioning device and a synchronized IMU. Ground truth labels for the car speed were calculated using the position measurements obtained at the high rate of 50 [Hz]. A DNN architecture with long short-term memory layers is proposed to enable high-frequency speed estimation that accounts for previous inputs history and the nonlinear relation between speed, acceleration, and angular velocity. A simplified aided dead reckoning localization scheme is formulated to assess the trained model which provides the speed pseudo-measurement. The trained model is shown to substantially improve the position accuracy during a 4 minutes drive without the use of GNSS position updates.

Policy Gradient Method For Robust Reinforcement Learning

This paper develops the first policy gradient method with global optimality guarantee and complexity analysis for robust reinforcement learning under model mismatch. Robust reinforcement learning is to learn a policy robust to model mismatch between simulator and real environment. We first develop the robust policy (sub-)gradient, which is applicable for any differentiable parametric policy class. We show that the proposed robust policy gradient method converges to the global optimum asymptotically under direct policy parameterization. We further develop a smoothed robust policy gradient method and show that to achieve an $\epsilon$-global optimum, the complexity is $\mathcal O(\epsilon^{-3})$. We then extend our methodology to the general model-free setting and design the robust actor-critic method with differentiable parametric policy class and value function. We further characterize its asymptotic convergence and sample complexity under the tabular setting. Finally, we provide simulation results to demonstrate the robustness of our methods.

Reductive MDPs: A Perspective Beyond Temporal Horizons

Solving general Markov decision processes (MDPs) is a computationally hard problem. Solving finite-horizon MDPs, on the other hand, is highly tractable with well known polynomial-time algorithms. What drives this extreme disparity, and do problems exist that lie between these diametrically opposed complexities? In this paper we identify and analyse a sub-class of stochastic shortest path problems (SSPs) for general state-action spaces whose dynamics satisfy a particular drift condition. This construction generalises the traditional, temporal notion of a horizon via decreasing reachability: a property called reductivity. It is shown that optimal policies can be recovered in polynomial-time for reductive SSPs -- via an extension of backwards induction -- with an efficient analogue in reductive MDPs. The practical considerations of the proposed approach are discussed, and numerical verification provided on a canonical optimal liquidation problem.

Sobolev Acceleration and Statistical Optimality for Learning Elliptic Equations via Gradient Descent

In this paper, we study the statistical limits in terms of Sobolev norms of gradient descent for solving inverse problem from randomly sampled noisy observations using a general class of objective functions. Our class of objective functions includes Sobolev training for kernel regression, Deep Ritz Methods (DRM), and Physics Informed Neural Networks (PINN) for solving elliptic partial differential equations (PDEs) as special cases. We consider a potentially infinite-dimensional parameterization of our model using a suitable Reproducing Kernel Hilbert Space and a continuous parameterization of problem hardness through the definition of kernel integral operators. We prove that gradient descent over this objective function can also achieve statistical optimality and the optimal number of passes over the data increases with sample size. Based on our theory, we explain an implicit acceleration of using a Sobolev norm as the objective function for training, inferring that the optimal number of epochs of DRM becomes larger than the number of PINN when both the data size and the hardness of tasks increase, although both DRM and PINN can achieve statistical optimality.

Analyzing Lottery Ticket Hypothesis from PAC-Bayesian Theory Perspective

The lottery ticket hypothesis (LTH) has attracted attention because it can explain why over-parameterized models often show high generalization ability. It is known that when we use iterative magnitude pruning (IMP), which is an algorithm to find sparse networks with high generalization ability that can be trained from the initial weights independently, called winning tickets, the initial large learning rate does not work well in deep neural networks such as ResNet. However, since the initial large learning rate generally helps the optimizer to converge to flatter minima, we hypothesize that the winning tickets have relatively sharp minima, which is considered a disadvantage in terms of generalization ability. In this paper, we confirm this hypothesis and show that the PAC-Bayesian theory can provide an explicit understanding of the relationship between LTH and generalization behavior. On the basis of our experimental findings that flatness is useful for improving accuracy and robustness to label noise and that the distance from the initial weights is deeply involved in winning tickets, we offer the PAC-Bayes bound using a spike-and-slab distribution to analyze winning tickets. Finally, we revisit existing algorithms for finding winning tickets from a PAC-Bayesian perspective and provide new insights into these methods.

cMelGAN: An Efficient Conditional Generative Model Based on Mel Spectrograms

Analysing music in the field of machine learning is a very difficult problem with numerous constraints to consider. The nature of audio data, with its very high dimensionality and widely varying scales of structure, is one of the primary reasons why it is so difficult to model. There are many applications of machine learning in music, like the classifying the mood of a piece of music, conditional music generation, or popularity prediction. The goal for this project was to develop a genre-conditional generative model of music based on Mel spectrograms and evaluate its performance by comparing it to existing generative music models that use note-based representations. We initially implemented an autoregressive, RNN-based generative model called MelNet . However, due to its slow speed and low fidelity output, we decided to create a new, fully convolutional architecture that is based on the MelGAN [4] and conditional GAN architectures, called cMelGAN.

Developing patient-driven artificial intelligence based on personal rankings of care decision making steps

We propose and experimentally motivate a new methodology to support decision-making processes in healthcare with artificial intelligence based on personal rankings of care decision making steps that can be identified with our methodology, questionnaire data and its statistical patterns. Our longitudinal quantitative cross-sectional three-stage study gathered self-ratings for 437 expression statements concerning healthcare situations on Likert scales in respect to "the need for help", "the advancement of health", "the hopefulness", "the indication of compassion" and "the health condition", and 45 answers about the person's demographics, health and wellbeing, also the duration of giving answers. Online respondents between 1 June 2020 and 29 June 2021 were recruited from Finnish patient and disabled people's organizations, other health-related organizations and professionals, and educational institutions (n=1075). With Kruskal-Wallis test, Wilcoxon rank-sum test (i.e., Mann-Whitney U test), Wilcoxon rank-sum pairwise test, Welch's t test and one-way analysis of variance (ANOVA) between groups test we identified statistically significant differences of ratings and their durations for each expression statement in respect to respondent groupings based on the answer values of each background question. Frequencies of the later reordering of rating rankings showed dependencies with ratings given earlier in respect to various interpretation task entities, interpretation dimensions and respondent groupings. Our methodology, questionnaire data and its statistical patterns enable analyzing with self-rated expression statements the representations of decision making steps in healthcare situations and their chaining, agglomeration and branching in knowledge entities of personalized care paths. Our results support building artificial intelligence solutions to address the patient's needs concerning care.

Parameter Adaptation for Joint Distribution Shifts

While different methods exist to tackle distinct types of distribution shift, such as label shift (in the form of adversarial attacks) or domain shift, tackling the joint shift setting is still an open problem. Through the study of a joint distribution shift manifesting both adversarial and domain-specific perturbations, we not only show that a joint shift worsens model performance compared to their individual shifts, but that the use of a similar domain worsens performance than a dissimilar domain. To curb the performance drop, we study the use of perturbation sets motivated by input and parameter space bounds, and adopt a meta learning strategy (hypernetworks) to model parameters w.r.t. test-time inputs to recover performance.

Generalization Bounds on Multi-Kernel Learning with Mixed Datasets

This paper presents novel generalization bounds for the multi-kernel learning problem. Motivated by applications in sensor networks, we assume that the dataset is mixed where each sample is taken from a finite pool of Markov chains. Our bounds for learning kernels admit $O(\sqrt{\log m})$ dependency on the number of base kernels and $O(1/\sqrt{n})$ dependency on the number of training samples. However, some $O(1/\sqrt{n})$ terms are added to compensate for the dependency among samples compared with existing generalization bounds for multi-kernel learning with i.i.d. datasets.

COIN: Communication-Aware In-Memory Acceleration for Graph Convolutional Networks

Graph convolutional networks (GCNs) have shown remarkable learning capabilities when processing graph-structured data found inherently in many application areas. GCNs distribute the outputs of neural networks embedded in each vertex over multiple iterations to take advantage of the relations captured by the underlying graphs. Consequently, they incur a significant amount of computation and irregular communication overheads, which call for GCN-specific hardware accelerators. To this end, this paper presents a communication-aware in-memory computing architecture (COIN) for GCN hardware acceleration. Besides accelerating the computation using custom compute elements (CE) and in-memory computing, COIN aims at minimizing the intra- and inter-CE communication in GCN operations to optimize the performance and energy efficiency. Experimental evaluations with widely used datasets show up to 105x improvement in energy consumption compared to state-of-the-art GCN accelerator.

3DLinker: An E(3) Equivariant Variational Autoencoder for Molecular Linker Design

Deep learning has achieved tremendous success in designing novel chemical compounds with desirable pharmaceutical properties. In this work, we focus on a new type of drug design problem -- generating a small "linker" to physically attach two independent molecules with their distinct functions. The main computational challenges include: 1) the generation of linkers is conditional on the two given molecules, in contrast to generating full molecules from scratch in previous works; 2) linkers heavily depend on the anchor atoms of the two molecules to be connected, which are not known beforehand; 3) 3D structures and orientations of the molecules need to be considered to avoid atom clashes, for which equivariance to E(3) group are necessary. To address these problems, we propose a conditional generative model, named 3DLinker, which is able to predict anchor atoms and jointly generate linker graphs and their 3D structures based on an E(3) equivariant graph variational autoencoder. So far as we know, there are no previous models that could achieve this task. We compare our model with multiple conditional generative models modified from other molecular design tasks and find that our model has a significantly higher rate in recovering molecular graphs, and more importantly, accurately predicting the 3D coordinates of all the atoms.

Finding Global Homophily in Graph Neural Networks When Meeting Heterophily

We investigate graph neural networks on graphs with heterophily. Some existing methods amplify a node's neighborhood with multi-hop neighbors to include more nodes with homophily. However, it is a significant challenge to set personalized neighborhood sizes for different nodes. Further, for other homophilous nodes excluded in the neighborhood, they are ignored for information aggregation. To address these problems, we propose two models GloGNN and GloGNN++, which generate a node's embedding by aggregating information from global nodes in the graph. In each layer, both models learn a coefficient matrix to capture the correlations between nodes, based on which neighborhood aggregation is performed. The coefficient matrix allows signed values and is derived from an optimization problem that has a closed-form solution. We further accelerate neighborhood aggregation and derive a linear time complexity. We theoretically explain the models' effectiveness by proving that both the coefficient matrix and the generated node embedding matrix have the desired grouping effect. We conduct extensive experiments to compare our models against 11 other competitors on 15 benchmark datasets in a wide range of domains, scales and graph heterophilies. Experimental results show that our methods achieve superior performance and are also very efficient.

Optimization of Decision Tree Evaluation Using SIMD Instructions

Decision forest (decision tree ensemble) is one of the most popular machine learning algorithms. To use large models on big data, like document scoring with learning-to-rank models, we need to evaluate these models efficiently. In this paper, we explore MatrixNet, the ancestor of the popular CatBoost library. Both libraries use the SSE instruction set for scoring on CPU. This paper investigates the opportunities given by the AVX instruction set to evaluate models more efficiently. We achieved 35% speedup on the binarization stage (nodes conditions comparison), and 20% speedup on the trees apply stage on the ranking model.

Posterior Probability Matters: Doubly-Adaptive Calibration for Neural Predictions in Online Advertising

Predicting user response probabilities is vital for ad ranking and bidding. We hope that predictive models can produce accurate probabilistic predictions that reflect true likelihoods. Calibration techniques aims to post-process model predictions to posterior probabilities. Field-level calibration -- which performs calibration w.r.t. to a specific field value -- is fine-grained and more practical. In this paper we propose a doubly-adaptive approach AdaCalib. It learns an isotonic function family to calibrate model predictions with the guidance of posterior statistics, and field-adaptive mechanisms are designed to ensure that the posterior is appropriate for the field value to be calibrated. Experiments verify that AdaCalib achieves significant improvement on calibration performance. It has been deployed online and beats previous approach.

A Computational Framework of Cortical Microcircuits Approximates Sign-concordant Random Backpropagation

Several recent studies attempt to address the biological implausibility of the well-known backpropagation (BP) method. While promising methods such as feedback alignment, direct feedback alignment, and their variants like sign-concordant feedback alignment tackle BP's weight transport problem, their validity remains controversial owing to a set of other unsolved issues. In this work, we answer the question of whether it is possible to realize random backpropagation solely based on mechanisms observed in neuroscience. We propose a hypothetical framework consisting of a new microcircuit architecture and its supporting Hebbian learning rules. Comprising three types of cells and two types of synaptic connectivity, the proposed microcircuit architecture computes and propagates error signals through local feedback connections and supports the training of multi-layered spiking neural networks with a globally defined spiking error function. We employ the Hebbian rule operating in local compartments to update synaptic weights and achieve supervised learning in a biologically plausible manner. Finally, we interpret the proposed framework from an optimization point of view and show its equivalence to sign-concordant feedback alignment. The proposed framework is benchmarked on several datasets including MNIST and CIFAR10, demonstrating promising BP-comparable accuracy.

A Note on the Chernoff Bound for Random Variables in the Unit Interval

The Chernoff bound is a well-known tool for obtaining a high probability bound on the expectation of a Bernoulli random variable in terms of its sample average. This bound is commonly used in statistical learning theory to upper bound the generalisation risk of a hypothesis in terms of its empirical risk on held-out data, for the case of a binary-valued loss function. However, the extension of this bound to the case of random variables taking values in the unit interval is less well known in the community. In this note we provide a proof of this extension for convenience and future reference.

Exploiting the Relationship Between Kendall's Rank Correlation and Cosine Similarity for Attribution Protection

Model attributions are important in deep neural networks as they aid practitioners in understanding the models, but recent studies reveal that attributions can be easily perturbed by adding imperceptible noise to the input. The non-differentiable Kendall's rank correlation is a key performance index for attribution protection. In this paper, we first show that the expected Kendall's rank correlation is positively correlated to cosine similarity and then indicate that the direction of attribution is the key to attribution robustness. Based on these findings, we explore the vector space of attribution to explain the shortcomings of attribution defense methods using $\ell_p$ norm and propose integrated gradient regularizer (IGR), which maximizes the cosine similarity between natural and perturbed attributions. Our analysis further exposes that IGR encourages neurons with the same activation states for natural samples and the corresponding perturbed samples, which is shown to induce robustness to gradient-based attribution methods. Our experiments on different models and datasets confirm our analysis on attribution protection and demonstrate a decent improvement in adversarial robustness.

Fairness via Explanation Quality: Evaluating Disparities in the Quality of Post hoc Explanations

As post hoc explanation methods are increasingly being leveraged to explain complex models in high-stakes settings, it becomes critical to ensure that the quality of the resulting explanations is consistently high across various population subgroups including the minority groups. For instance, it should not be the case that explanations associated with instances belonging to a particular gender subgroup (e.g., female) are less accurate than those associated with other genders. However, there is little to no research that assesses if there exist such group-based disparities in the quality of the explanations output by state-of-the-art explanation methods. In this work, we address the aforementioned gaps by initiating the study of identifying group-based disparities in explanation quality. To this end, we first outline the key properties which constitute explanation quality and where disparities can be particularly problematic. We then leverage these properties to propose a novel evaluation framework which can quantitatively measure disparities in the quality of explanations output by state-of-the-art methods. Using this framework, we carry out a rigorous empirical analysis to understand if and when group-based disparities in explanation quality arise. Our results indicate that such disparities are more likely to occur when the models being explained are complex and highly non-linear. In addition, we also observe that certain post hoc explanation methods (e.g., Integrated Gradients, SHAP) are more likely to exhibit the aforementioned disparities. To the best of our knowledge, this work is the first to highlight and study the problem of group-based disparities in explanation quality. In doing so, our work sheds light on previously unexplored ways in which explanation methods may introduce unfairness in real world decision making.

Classifiers are Better Experts for Controllable Text Generation

This paper proposes a simple method for controllable text generation based on weighting logits with a free-form classifier, namely CAIF sampling. Using an arbitrary text classifier, we adjust a small part of a language model's logits and guide text generation towards or away from classifier prediction. We experimented with toxicity avoidance and sentiment control tasks and showed that the proposed method significantly outperforms recent PPLM, GeDi, and DExperts on PPL and task accuracy metrics based on the external classifier of generated texts. In addition, compared to other approaches, it is easier to implement and tune and has significantly fewer restrictions and requirements.

Supervised Learning and Model Analysis with Compositional Data

The compositionality and sparsity of high-throughput sequencing data poses a challenge for regression and classification. However, in microbiome research in particular, conditional modeling is an essential tool to investigate relationships between phenotypes and the microbiome. Existing techniques are often inadequate: they either rely on extensions of the linear log-contrast model (which adjusts for compositionality, but is often unable to capture useful signals), or they are based on black-box machine learning methods (which may capture useful signals, but ignore compositionality in downstream analyses). We propose KernelBiome, a kernel-based nonparametric regression and classification framework for compositional data. It is tailored to sparse compositional data and is able to incorporate prior knowledge, such as phylogenetic structure. KernelBiome captures complex signals, including in the zero-structure, while automatically adapting model complexity. We demonstrate on par or improved predictive performance compared with state-of-the-art machine learning methods. Additionally, our framework provides two key advantages: (i) We propose two novel quantities to interpret contributions of individual components and prove that they consistently estimate average perturbation effects of the conditional mean, extending the interpretability of linear log-contrast models to nonparametric models. (ii) We show that the connection between kernels and distances aids interpretability and provides a data-driven embedding that can augment further analysis. Finally, we apply the KernelBiome framework to two public microbiome studies and illustrate the proposed model analysis. KernelBiome is available as an open-source Python package at https://github.com/shimenghuang/KernelBiome.

Textual Explanations and Critiques in Recommendation Systems

Artificial intelligence and machine learning algorithms have become ubiquitous. Although they offer a wide range of benefits, their adoption in decision-critical fields is limited by their lack of interpretability, particularly with textual data. Moreover, with more data available than ever before, it has become increasingly important to explain automated predictions. Generally, users find it difficult to understand the underlying computational processes and interact with the models, especially when the models fail to generate the outcomes or explanations, or both, correctly. This problem highlights the growing need for users to better understand the models' inner workings and gain control over their actions. This dissertation focuses on two fundamental challenges of addressing this need. The first involves explanation generation: inferring high-quality explanations from text documents in a scalable and data-driven manner. The second challenge consists in making explanations actionable, and we refer to it as critiquing. This dissertation examines two important applications in natural language processing and recommendation tasks. Overall, we demonstrate that interpretability does not come at the cost of reduced performance in two consequential applications. Our framework is applicable to other fields as well. This dissertation presents an effective means of closing the gap between promise and practice in artificial intelligence.

Discovering the Representation Bottleneck of Graph Neural Networks from Multi-order Interactions

Most graph neural networks (GNNs) rely on the message passing paradigm to propagate node features and build interactions. Recent works point out that different graph learning tasks require different ranges of interactions between nodes. To investigate the underlying mechanism, we explore the capacity of GNNs to capture pairwise interactions between nodes under contexts with different complexities, especially for their graph-level and node-level applications in scientific domains like biochemistry and physics. When formulating pairwise interactions, we study two standard graph construction methods in scientific domains, i.e., K-nearest neighbor (KNN) graphs and fully-connected (FC) graphs. Furthermore, we demonstrate that the inductive bias introduced by KNN-graphs and FC-graphs inhibits GNNs from learning the most informative order of interactions. Such a phenomenon is broadly shared by several GNNs for different graph learning tasks and prevents GNNs from reaching the global minimum loss, so we name it a representation bottleneck. To overcome that, we propose a novel graph rewiring approach based on the pairwise interaction strengths to adjust the reception fields of each node dynamically. Extensive experiments in molecular property prediction and dynamic system forecast prove the superiority of our method over state-of-the-art GNN baselines. Besides, this paper provides a reasonable explanation of why subgraphs play a vital role in determining graph properties. The code is available at https://github.com/smiles724/bottleneck.

Topic Modelling on Consumer Financial Protection Bureau Data: An Approach Using BERT Based Embeddings

Customers' reviews and comments are important for businesses to understand users' sentiment about the products and services. However, this data needs to be analyzed to assess the sentiment associated with topics/aspects to provide efficient customer assistance. LDA and LSA fail to capture the semantic relationship and are not specific to any domain. In this study, we evaluate BERTopic, a novel method that generates topics using sentence embeddings on Consumer Financial Protection Bureau (CFPB) data. Our work shows that BERTopic is flexible and yet provides meaningful and diverse topics compared to LDA and LSA. Furthermore, domain-specific pre-trained embeddings (FinBERT) yield even better topics. We evaluated the topics on coherence score (c_v) and UMass.

Not to Overfit or Underfit? A Study of Domain Generalization in Question Answering

Machine learning models are prone to overfitting their source (training) distributions, which is commonly believed to be why they falter in novel target domains. Here we examine the contrasting view that multi-source domain generalization (DG) is in fact a problem of mitigating source domain underfitting: models not adequately learning the signal in their multi-domain training data. Experiments on a reading comprehension DG benchmark show that as a model gradually learns its source domains better -- using known methods such as knowledge distillation from a larger model -- its zero-shot out-of-domain accuracy improves at an even faster rate. Improved source domain learning also demonstrates superior generalization over three popular domain-invariant learning methods that aim to counter overfitting.

Evaluating Independence and Conditional Independence Measures

Independence and Conditional Independence (CI) are two fundamental concepts in probability and statistics, which can be applied to solve many central problems of statistical inference. There are many existing independence and CI measures defined from diverse principles and concepts. In this paper, the 16 independence measures and 16 CI measures were reviewed and then evaluated with simulated and real data. For the independence measures, eight simulated data were generating from normal distribution, normal and Archimedean copula functions to compare the measures in bivariate or multivariate, linear or nonlinear settings. Two UCI dataset, including the heart disease data and the wine quality data, were used to test the power of the independence measures in real conditions. For the CI measures, two simulated data with normal distribution and Gumbel copula, and one real data (the Beijing air data) were utilized to test the CI measures in prespecified linear or nonlinear setting and real scenario. From the experimental results, we found that most of the measures work well on the simulated data by presenting the right monotonicity of the simulations. However, the independence and CI measures were differentiated on much complex real data respectively and only a few can be considered as working well with reference to domain knowledge. We also found that the measures tend to be separated into groups based on the similarity of the behaviors of them in each setting and in general. According to the experiments, we recommend CE as a good choice for both independence and CI measure. This is also due to its rigorous distribution-free definition and consistent nonparametric estimator.

Reliable Offline Model-based Optimization for Industrial Process Control

In the research area of offline model-based optimization, novel and promising methods are frequently developed. However, implementing such methods in real-world industrial systems such as production lines for process control is oftentimes a frustrating process. In this work, we address two important problems to extend the current success of offline model-based optimization to industrial process control problems: 1) how to learn a reliable dynamics model from offline data for industrial processes? 2) how to learn a reliable but not over-conservative control policy from offline data by utilizing existing model-based optimization algorithms? Specifically, we propose a dynamics model based on ensemble of conditional generative adversarial networks to achieve accurate reward calculation in industrial scenarios. Furthermore, we propose an epistemic-uncertainty-penalized reward evaluation function which can effectively avoid giving over-estimated rewards to out-of-distribution inputs during the learning/searching of the optimal control policy. We provide extensive experiments with the proposed method on two representative cases (a discrete control case and a continuous control case), showing that our method compares favorably to several baselines in offline policy learning for industrial process control.

Pocket2Mol: Efficient Molecular Sampling Based on 3D Protein Pockets

Deep generative models have achieved tremendous success in designing novel drug molecules in recent years. A new thread of works have shown the great potential in advancing the specificity and success rate of in silico drug design by considering the structure of protein pockets. This setting posts fundamental computational challenges in sampling new chemical compounds that could satisfy multiple geometrical constraints imposed by pockets. Previous sampling algorithms either sample in the graph space or only consider the 3D coordinates of atoms while ignoring other detailed chemical structures such as bond types and functional groups. To address the challenge, we develop Pocket2Mol, an E(3)-equivariant generative network composed of two modules: 1) a new graph neural network capturing both spatial and bonding relationships between atoms of the binding pockets and 2) a new efficient algorithm which samples new drug candidates conditioned on the pocket representations from a tractable distribution without relying on MCMC. Experimental results demonstrate that molecules sampled from Pocket2Mol achieve significantly better binding affinity and other drug properties such as druglikeness and synthetic accessibility.

Learning Shared Kernel Models: the Shared Kernel EM algorithm

Expectation maximisation (EM) is an unsupervised learning method for estimating the parameters of a finite mixture distribution. It works by introducing "hidden" or "latent" variables via Baum's auxiliary function $Q$ that allow the joint data likelihood to be expressed as a product of simple factors. The relevance of EM has increased since the introduction of the variational lower bound (VLB): the VLB differs from Baum's auxiliary function only by the entropy of the PDF of the latent variables $Z$. We first present a rederivation of the standard EM algorithm using data association ideas from the field of multiple target tracking, using $K$-valued scalar data association hypotheses rather than the usual binary indicator vectors. The same method is then applied to a little known but much more general type of supervised EM algorithm for shared kernel models, related to probabilistic radial basis function networks. We address a number of shortcomings in the derivations that have been published previously in this area. In particular, we give theoretically rigorous derivations of (i) the complete data likelihood; (ii) Baum's auxiliary function (the E-step) and (iii) the maximisation (M-step) in the case of Gaussian shared kernel models. The subsequent algorithm, called shared kernel EM (SKEM), is then applied to a digit recognition problem using a novel 7-segment digit representation. Variants of the algorithm that use different numbers of features and different EM algorithm dimensions are compared in terms of mean accuracy and mean IoU. A simplified classifier is proposed that decomposes the joint data PDF as a product of lower order PDFs over non-overlapping subsets of variables. The effect of different numbers of assumed mixture components $K$ is also investigated. High-level source code for the data generation and SKEM algorithm is provided.

FreeMatch: Self-adaptive Thresholding for Semi-supervised Learning

Pseudo labeling and consistency regularization approaches based on confidence thresholding have made great progress in semi-supervised learning (SSL). However, we argue that existing methods might fail to adopt suitable thresholds since they either use a pre-defined / fixed threshold or an ad-hoc threshold adjusting scheme, resulting in inferior performance and slow convergence. We first analyze a motivating example to achieve some intuitions on the relationship between the desirable threshold and model's learning status. Based on the analysis, we hence propose FreeMatch to define and adjust the confidence threshold in a self-adaptive manner according to the model's learning status. We further introduce a self-adaptive class fairness regularization penalty that encourages the model to produce diverse predictions during the early stages of training. Extensive experimental results indicate the superiority of FreeMatch especially when the labeled data are extremely rare. FreeMatch achieves 5.78%, 13.59%, and 1.28% error rate reduction over the latest state-of-the-art method FlexMatch on CIFAR-10 with 1 label per class, STL-10 with 4 labels per class, and ImageNet with 100 labels per class, respectively.

Combating COVID-19 using Generative Adversarial Networks and Artificial Intelligence for Medical Images: A Scoping Review

This review presents a comprehensive study on the role of GANs in addressing the challenges related to COVID-19 data scarcity and diagnosis. It is the first review that summarizes the different GANs methods and the lungs images datasets for COVID-19. It attempts to answer the questions related to applications of GANs, popular GAN architectures, frequently used image modalities, and the availability of source code. This review included 57 full-text studies that reported the use of GANs for different applications in COVID-19 lungs images data. Most of the studies (n=42) used GANs for data augmentation to enhance the performance of AI techniques for COVID-19 diagnosis. Other popular applications of GANs were segmentation of lungs and super-resolution of the lungs images. The cycleGAN and the conditional GAN were the most commonly used architectures used in nine studies each. 29 studies used chest X-Ray images while 21 studies used CT images for the training of GANs. For majority of the studies (n=47), the experiments were done and results were reported using publicly available data. A secondary evaluation of the results by radiologists/clinicians was reported by only two studies. Conclusion: Studies have shown that GANs have great potential to address the data scarcity challenge for lungs images of COVID-19. Data synthesized with GANs have been helpful to improve the training of the Convolutional Neural Network (CNN) models trained for the diagnosis of COVID-19. Besides, GANs have also contributed to enhancing the CNNs performance through the super-resolution of the images and segmentation. This review also identified key limitations of the potential transformation of GANs based methods in clinical applications.

Clinical outcome prediction under hypothetical interventions -- a representation learning framework for counterfactual reasoning

Most machine learning (ML) models are developed for prediction only; offering no option for causal interpretation of their predictions or parameters/properties. This can hamper the health systems' ability to employ ML models in clinical decision-making processes, where the need and desire for predicting outcomes under hypothetical investigations (i.e., counterfactual reasoning/explanation) is high. In this research, we introduce a new representation learning framework (i.e., partial concept bottleneck), which considers the provision of counterfactual explanations as an embedded property of the risk model. Despite architectural changes necessary for jointly optimising for prediction accuracy and counterfactual reasoning, the accuracy of our approach is comparable to prediction-only models. Our results suggest that our proposed framework has the potential to help researchers and clinicians improve personalised care (e.g., by investigating the hypothetical differential effects of interventions)

RoMFAC: A Robust Mean-Field Actor-Critic Reinforcement Learning against Adversarial Perturbations on States

Deep reinforcement learning methods for multi-agent systems make optimal decisions dependent on states observed by agents, but a little uncertainty on the observations can possibly mislead agents into taking wrong actions. The mean-field actor-critic reinforcement learning (MFAC) is very famous in the multi-agent field since it can effectively handle the scalability problem. However, this paper finds that it is also sensitive to state perturbations which can significantly degrade the team rewards. This paper proposes a robust learning framework for MFAC called RoMFAC that has two innovations: 1) a new objective function of training actors, composed of a \emph{policy gradient function} that is related to the expected cumulative discount reward on sampled clean states and an \emph{action loss function} that represents the difference between actions taken on clean and adversarial states; and 2) a repetitive regularization of the action loss that ensures the trained actors obtain a good performance. Furthermore, we prove that the proposed action loss function is convergent. Experiments show that RoMFAC is robust against adversarial perturbations while maintaining its good performance in environments without perturbations.

Sample-Efficient Learning of Correlated Equilibria in Extensive-Form Games

Imperfect-Information Extensive-Form Games (IIEFGs) is a prevalent model for real-world games involving imperfect information and sequential plays. The Extensive-Form Correlated Equilibrium (EFCE) has been proposed as a natural solution concept for multi-player general-sum IIEFGs. However, existing algorithms for finding an EFCE require full feedback from the game, and it remains open how to efficiently learn the EFCE in the more challenging bandit feedback setting where the game can only be learned by observations from repeated playing. This paper presents the first sample-efficient algorithm for learning the EFCE from bandit feedback. We begin by proposing $K$-EFCE -- a more generalized definition that allows players to observe and deviate from the recommended actions for $K$ times. The $K$-EFCE includes the EFCE as a special case at $K=1$, and is an increasingly stricter notion of equilibrium as $K$ increases. We then design an uncoupled no-regret algorithm that finds an $\varepsilon$-approximate $K$-EFCE within $\widetilde{\mathcal{O}}(\max_{i}X_iA_i^{K}/\varepsilon^2)$ iterations in the full feedback setting, where $X_i$ and $A_i$ are the number of information sets and actions for the $i$-th player. Our algorithm works by minimizing a wide-range regret at each information set that takes into account all possible recommendation histories. Finally, we design a sample-based variant of our algorithm that learns an $\varepsilon$-approximate $K$-EFCE within $\widetilde{\mathcal{O}}(\max_{i}X_iA_i^{K+1}/\varepsilon^2)$ episodes of play in the bandit feedback setting. When specialized to $K=1$, this gives the first sample-efficient algorithm for learning EFCE from bandit feedback.

Online Nonsubmodular Minimization with Delayed Costs: From Full Information to Bandit Feedback

Motivated by applications to online learning in sparse estimation and Bayesian optimization, we consider the problem of online unconstrained nonsubmodular minimization with delayed costs in both full information and bandit feedback settings. In contrast to previous works on online unconstrained submodular minimization, we focus on a class of nonsubmodular functions with special structure, and prove regret guarantees for several variants of the online and approximate online bandit gradient descent algorithms in static and delayed scenarios. We derive bounds for the agent's regret in the full information and bandit feedback setting, even if the delay between choosing a decision and receiving the incurred cost is unbounded. Key to our approach is the notion of $(\alpha, \beta)$-regret and the extension of the generic convex relaxation model from~\citet{El-2020-Optimal}, the analysis of which is of independent interest. We conduct and showcase several simulation studies to demonstrate the efficacy of our algorithms.

Federated learning for LEO constellations via inter-HAP links

Low Earth Obit (LEO) satellite constellations have seen a sharp increase of deployment in recent years, due to their distinctive capabilities of providing broadband Internet access and enabling global data acquisition as well as large-scale AI applications. To apply machine learning (ML) in such applications, the traditional way of downloading satellite data such as imagery to a ground station (GS) and then training a model in a centralized manner, is not desirable because of the limited bandwidth, intermittent connectivity between satellites and the GS, and privacy concerns on transmitting raw data. Federated Learning (FL) as an emerging communication and computing paradigm provides a potentially supreme solution to this problem. However, we show that existing FL solutions do not fit well in such LEO constellation scenarios because of significant challenges such as excessive convergence delay and unreliable wireless channels. To this end, we propose to introduce high-altitude platforms (HAPs) as distributed parameter servers (PSs) and propose a synchronous FL algorithm, FedHAP, to accomplish model training in an efficient manner via inter-satellite collaboration. To accelerate convergence, we also propose a layered communication scheme between satellites and HAPs that FedHAP leverages. Our simulations demonstrate that FedHAP attains model convergence in much fewer communication rounds than benchmarks, cutting the training time substantially from several days down to a few hours with the same level of resulting accuracy.

Towards a Comprehensive Solution for a Vision-based Digitized Neurological Examination

The ability to use digitally recorded and quantified neurological exam information is important to help healthcare systems deliver better care, in-person and via telehealth, as they compensate for a growing shortage of neurologists. Current neurological digital biomarker pipelines, however, are narrowed down to a specific neurological exam component or applied for assessing specific conditions. In this paper, we propose an accessible vision-based exam and documentation solution called Digitized Neurological Examination (DNE) to expand exam biomarker recording options and clinical applications using a smartphone/tablet. Through our DNE software, healthcare providers in clinical settings and people at home are enabled to video capture an examination while performing instructed neurological tests, including finger tapping, finger to finger, forearm roll, and stand-up and walk. Our modular design of the DNE software supports integrations of additional tests. The DNE extracts from the recorded examinations the 2D/3D human-body pose and quantifies kinematic and spatio-temporal features. The features are clinically relevant and allow clinicians to document and observe the quantified movements and the changes of these metrics over time. A web server and a user interface for recordings viewing and feature visualizations are available. DNE was evaluated on a collected dataset of 21 subjects containing normal and simulated-impaired movements. The overall accuracy of DNE is demonstrated by classifying the recorded movements using various machine learning models. Our tests show an accuracy beyond 90% for upper-limb tests and 80% for the stand-up and walk tests.

A comparison of PINN approaches for drift-diffusion equations on metric graphs

In this paper we focus on comparing machine learning approaches for quantum graphs, which are metric graphs, i.e., graphs with dedicated edge lengths, and an associated differential operator. In our case the differential equation is a drift-diffusion model. Computational methods for quantum graphs require a careful discretization of the differential operator that also incorporates the node conditions, in our case Kirchhoff-Neumann conditions. Traditional numerical schemes are rather mature but have to be tailored manually when the differential equation becomes the constraint in an optimization problem. Recently, physics informed neural networks (PINNs) have emerged as a versatile tool for the solution of partial differential equations from a range of applications. They offer flexibility to solve parameter identification or optimization problems by only slightly changing the problem formulation used for the forward simulation. We compare several PINN approaches for solving the drift-diffusion on the metric graph.

Fair Bayes-Optimal Classifiers Under Predictive Parity

Increasing concerns about disparate effects of AI have motivated a great deal of work on fair machine learning. Existing works mainly focus on independence- and separation-based measures (e.g., demographic parity, equality of opportunity, equalized odds), while sufficiency-based measures such as predictive parity are much less studied. This paper considers predictive parity, which requires equalizing the probability of success given a positive prediction among different protected groups. We prove that, if the overall performances of different groups vary only moderately, all fair Bayes-optimal classifiers under predictive parity are group-wise thresholding rules. Perhaps surprisingly, this may not hold if group performance levels vary widely; in this case we find that predictive parity among protected groups may lead to within-group unfairness. We then propose an algorithm we call FairBayes-DPP, aiming to ensure predictive parity when our condition is satisfied. FairBayes-DPP is an adaptive thresholding algorithm that aims to achieve predictive parity, while also seeking to maximize test accuracy. We provide supporting experiments conducted on synthetic and empirical data.

Sparsity-Aware Robust Normalized Subband Adaptive Filtering algorithms based on Alternating Optimization

This paper proposes a unified sparsity-aware robust normalized subband adaptive filtering (SA-RNSAF) algorithm for identification of sparse systems under impulsive noise. The proposed SA-RNSAF algorithm generalizes different algorithms by defining the robust criterion and sparsity-aware penalty. Furthermore, by alternating optimization of the parameters (AOP) of the algorithm, including the step-size and the sparsity penalty weight, we develop the AOP-SA-RNSAF algorithm, which not only exhibits fast convergence but also obtains low steady-state misadjustment for sparse systems. Simulations in various noise scenarios have verified that the proposed AOP-SA-RNSAF algorithm outperforms existing techniques.

Proxyless Neural Architecture Adaptation for Supervised Learning and Self-Supervised Learning

Recently, Neural Architecture Search (NAS) methods have been introduced and show impressive performance on many benchmarks. Among those NAS studies, Neural Architecture Transformer (NAT) aims to adapt the given neural architecture to improve performance while maintaining computational costs. However, NAT lacks reproducibility and it requires an additional architecture adaptation process before network weight training. In this paper, we propose proxyless neural architecture adaptation that is reproducible and efficient. Our method can be applied to both supervised learning and self-supervised learning. The proposed method shows stable performance on various architectures. Extensive reproducibility experiments on two datasets, i.e., CIFAR-10 and Tiny Imagenet, present that the proposed method definitely outperforms NAT and is applicable to other models and datasets.

Evaluating Generalizability of Fine-Tuned Models for Fake News Detection

The Covid-19 pandemic has caused a dramatic and parallel rise in dangerous misinformation, denoted an `infodemic' by the CDC and WHO. Misinformation tied to the Covid-19 infodemic changes continuously; this can lead to performance degradation of fine-tuned models due to concept drift. Degredation can be mitigated if models generalize well-enough to capture some cyclical aspects of drifted data. In this paper, we explore generalizability of pre-trained and fine-tuned fake news detectors across 9 fake news datasets. We show that existing models often overfit on their training dataset and have poor performance on unseen data. However, on some subsets of unseen data that overlap with training data, models have higher accuracy. Based on this observation, we also present KMeans-Proxy, a fast and effective method based on K-Means clustering for quickly identifying these overlapping subsets of unseen data. KMeans-Proxy improves generalizability on unseen fake news datasets by 0.1-0.2 f1-points across datasets. We present both our generalizability experiments as well as KMeans-Proxy to further research in tackling the fake news problem.

Interpretable Stochastic Model Predictive Control using Distributional Reinforced Estimation for Quadrotor Tracking Systems

This paper presents a novel trajectory tracker for autonomous quadrotor navigation in dynamic and complex environments. The proposed framework integrates a distributional Reinforcement Learning (RL) estimator for unknown aerodynamic effects into a Stochastic Model Predictive Controller (SMPC) for trajectory tracking. Aerodynamic effects derived from drag forces and moment variations are difficult to model directly and accurately. Most current quadrotor tracking systems therefore treat them as simple `disturbances' in conventional control approaches. We propose Quantile-approximation-based Distributional Reinforced-disturbance-estimator, an aerodynamic disturbance estimator, to accurately identify disturbances, i.e., uncertainties between the true and estimated values of aerodynamic effects. Simplified Affine Disturbance Feedback is employed for control parameterization to guarantee convexity, which we then integrate with a SMPC to achieve sufficient and non-conservative control signals. We demonstrate our system to improve the cumulative tracking errors by at least 66% with unknown and diverse aerodynamic forces compared with recent state-of-the-art. Concerning traditional Reinforcement Learning's non-interpretability, we provide convergence and stability guarantees of Distributional RL and SMPC, respectively, with non-zero mean disturbances.

Trajectory Inference via Mean-field Langevin in Path Space

Trajectory inference aims at recovering the dynamics of a population from snapshots of its temporal marginals. To solve this task, a min-entropy estimator relative to the Wiener measure in path space was introduced by Lavenant et al. arXiv:2102.09204, and shown to consistently recover the dynamics of a large class of drift-diffusion processes from the solution of an infinite dimensional convex optimization problem. In this paper, we introduce a grid-free algorithm to compute this estimator. Our method consists in a family of point clouds (one per snapshot) coupled via Schr\"odinger bridges which evolve with noisy gradient descent. We study the mean-field limit of the dynamics and prove its global convergence to the desired estimator. Overall, this leads to an inference method with end-to-end theoretical guarantees that solves an interpretable model for trajectory inference. We also present how to adapt the method to deal with mass variations, a useful extension when dealing with single cell RNA-sequencing data where cells can branch and die.

BackLink: Supervised Local Training with Backward Links

Empowered by the backpropagation (BP) algorithm, deep neural networks have dominated the race in solving various cognitive tasks. The restricted training pattern in the standard BP requires end-to-end error propagation, causing large memory cost and prohibiting model parallelization. Existing local training methods aim to resolve the training obstacle by completely cutting off the backward path between modules and isolating their gradients to reduce memory cost and accelerate the training process. These methods prevent errors from flowing between modules and hence information exchange, resulting in inferior performance. This work proposes a novel local training algorithm, BackLink, which introduces inter-module backward dependency and allows errors to flow between modules. The algorithm facilitates information to flow backward along with the network. To preserve the computational advantage of local training, BackLink restricts the error propagation length within the module. Extensive experiments performed in various deep convolutional neural networks demonstrate that our method consistently improves the classification performance of local training algorithms over other methods. For example, in ResNet32 with 16 local modules, our method surpasses the conventional greedy local training method by 4.00\% and a recent work by 1.83\% in accuracy on CIFAR10, respectively. Analysis of computational costs reveals that small overheads are incurred in GPU memory costs and runtime on multiple GPUs. Our method can lead up to a 79\% reduction in memory cost and 52\% in simulation runtime in ResNet110 compared to the standard BP. Therefore, our method could create new opportunities for improving training algorithms towards better efficiency and biological plausibility.

Breaking with Fixed Set Pathology Recognition through Report-Guided Contrastive Training

When reading images, radiologists generate text reports describing the findings therein. Current state-of-the-art computer-aided diagnosis tools utilize a fixed set of predefined categories automatically extracted from these medical reports for training. This form of supervision limits the potential usage of models as they are unable to pick up on anomalies outside of their predefined set, thus, making it a necessity to retrain the classifier with additional data when faced with novel classes. In contrast, we investigate direct text supervision to break away from this closed set assumption. By doing so, we avoid noisy label extraction via text classifiers and incorporate more contextual information. We employ a contrastive global-local dual-encoder architecture to learn concepts directly from unstructured medical reports while maintaining its ability to perform free form classification. We investigate relevant properties of open set recognition for radiological data and propose a method to employ currently weakly annotated data into training. We evaluate our approach on the large-scale chest X-Ray datasets MIMIC-CXR, CheXpert, and ChestX-Ray14 for disease classification. We show that despite using unstructured medical report supervision, we perform on par with direct label supervision through a sophisticated inference setting.

Classification of Astronomical Bodies by Efficient Layer Fine-Tuning of Deep Neural Networks

The SDSS-IV dataset contains information about various astronomical bodies such as Galaxies, Stars, and Quasars captured by observatories. Inspired by our work on deep multimodal learning, which utilized transfer learning to classify the SDSS-IV dataset, we further extended our research in the fine tuning of these architectures to study the effect in the classification scenario. Architectures such as Resnet-50, DenseNet-121 VGG-16, Xception, EfficientNetB2, MobileNetV2 and NasnetMobile have been built using layer wise fine tuning at different levels. Our findings suggest that freezing all layers with Imagenet weights and adding a final trainable layer may not be the optimal solution. Further, baseline models and models that have higher number of trainable layers performed similarly in certain architectures. Model need to be fine tuned at different levels and a specific training ratio is required for a model to be termed ideal. Different architectures had different responses to the change in the number of trainable layers w.r.t accuracies. While models such as DenseNet-121, Xception, EfficientNetB2 achieved peak accuracies that were relatively consistent with near perfect training curves, models such as Resnet-50,VGG-16, MobileNetV2 and NasnetMobile had lower, delayed peak accuracies with poorly fitting training curves. It was also found that though mobile neural networks have lesser parameters and model size, they may not always be ideal for deployment on a low computational device as they had consistently lower validation accuracies. Customized evaluation metrics such as Tuning Parameter Ratio and Tuning Layer Ratio are used for model evaluation.

Revisiting Facial Key Point Detection: An Efficient Approach Using Deep Neural Networks

Facial landmark detection is a widely researched field of deep learning as this has a wide range of applications in many fields. These key points are distinguishing characteristic points on the face, such as the eyes center, the eye's inner and outer corners, the mouth center, and the nose tip from which human emotions and intent can be explained. The focus of our work has been evaluating transfer learning models such as MobileNetV2 and NasNetMobile, including custom CNN architectures. The objective of the research has been to develop efficient deep learning models in terms of model size, parameters, and inference time and to study the effect of augmentation imputation and fine-tuning on these models. It was found that while augmentation techniques produced lower RMSE scores than imputation techniques, they did not affect the inference time. MobileNetV2 architecture produced the lowest RMSE and inference time. Moreover, our results indicate that manually optimized CNN architectures performed similarly to Auto Keras tuned architecture. However, manually optimized architectures yielded better inference time and training curves.

Efficient Deep Learning Methods for Identification of Defective Casting Products

Quality inspection has become crucial in any large-scale manufacturing industry recently. In order to reduce human error, it has become imperative to use efficient and low computational AI algorithms to identify such defective products. In this paper, we have compared and contrasted various pre-trained and custom-built architectures using model size, performance and CPU latency in the detection of defective casting products. Our results show that custom architectures are efficient than pre-trained mobile architectures. Moreover, custom models perform 6 to 9 times faster than lightweight models such as MobileNetV2 and NasNet. The number of training parameters and the model size of the custom architectures is significantly lower (~386 times & ~119 times respectively) than the best performing models such as MobileNetV2 and NasNet. Augmentation experimentations have also been carried out on the custom architectures to make the models more robust and generalizable. Our work sheds light on the efficiency of these custom-built architectures for deployment on Edge and IoT devices and that transfer learning models may not always be ideal. Instead, they should be specific to the kind of dataset and the classification problem at hand.

Practical Insights of Repairing Model Problems on Image Classification

Additional training of a deep learning model can cause negative effects on the results, turning an initially positive sample into a negative one (degradation). Such degradation is possible in real-world use cases due to the diversity of sample characteristics. That is, a set of samples is a mixture of critical ones which should not be missed and less important ones. Therefore, we cannot understand the performance by accuracy alone. While existing research aims to prevent a model degradation, insights into the related methods are needed to grasp their benefits and limitations. In this talk, we will present implications derived from a comparison of methods for reducing degradation. Especially, we formulated use cases for industrial settings in terms of arrangements of a data set. The results imply that a practitioner should care about better method continuously considering dataset availability and life cycle of an AI system because of a trade-off between accuracy and preventing degradation.

SystemMatch: optimizing preclinical drug models to human clinical outcomes via generative latent-space matching

Translating the relevance of preclinical models ($\textit{in vitro}$, animal models, or organoids) to their relevance in humans presents an important challenge during drug development. The rising abundance of single-cell genomic data from human tumors and tissue offers a new opportunity to optimize model systems by their similarity to targeted human cell types in disease. In this work, we introduce SystemMatch to assess the fit of preclinical model systems to an $\textit{in sapiens}$ target population and to recommend experimental changes to further optimize these systems. We demonstrate this through an application to developing $\textit{in vitro}$ systems to model human tumor-derived suppressive macrophages. We show with held-out $\textit{in vivo}$ controls that our pipeline successfully ranks macrophage subpopulations by their biological similarity to the target population, and apply this analysis to rank a series of 18 $\textit{in vitro}$ macrophage systems perturbed with a variety of cytokine stimulations. We extend this analysis to predict the behavior of 66 $\textit{in silico}$ model systems generated using a perturbational autoencoder and apply a $k$-medoids approach to recommend a subset of these model systems for further experimental development in order to fully explore the space of possible perturbations. Through this use case, we demonstrate a novel approach to model system development to generate a system more similar to human biology.

Unsupervised Abnormal Traffic Detection through Topological Flow Analysis

Cyberthreats are a permanent concern in our modern technological world. In the recent years, sophisticated traffic analysis techniques and anomaly detection (AD) algorithms have been employed to face the more and more subversive adversarial attacks. A malicious intrusion, defined as an invasive action intending to illegally exploit private resources, manifests through unusual data traffic and/or abnormal connectivity pattern. Despite the plethora of statistical or signature-based detectors currently provided in the literature, the topological connectivity component of a malicious flow is less exploited. Furthermore, a great proportion of the existing statistical intrusion detectors are based on supervised learning, that relies on labeled data. By viewing network flows as weighted directed interactions between a pair of nodes, in this paper we present a simple method that facilitate the use of connectivity graph features in unsupervised anomaly detection algorithms. We test our methodology on real network traffic datasets and observe several improvements over standard AD.

Robust Regularized Low-Rank Matrix Models for Regression and Classification

While matrix variate regression models have been studied in many existing works, classical statistical and computational methods for the analysis of the regression coefficient estimation are highly affected by high dimensional and noisy matrix-valued predictors. To address these issues, this paper proposes a framework of matrix variate regression models based on a rank constraint, vector regularization (e.g., sparsity), and a general loss function with three special cases considered: ordinary matrix regression, robust matrix regression, and matrix logistic regression. We also propose an alternating projected gradient descent algorithm. Based on analyzing our objective functions on manifolds with bounded curvature, we show that the algorithm is guaranteed to converge, all accumulation points of the iterates have estimation errors in the order of $O(1/\sqrt{n})$ asymptotically and substantially attaining the minimax rate. Our theoretical analysis can be applied to general optimization problems on manifolds with bounded curvature and can be considered an important technical contribution to this work. We validate the proposed method through simulation studies and real image data examples.

A Comprehensive Survey on Model Quantization for Deep Neural Networks

Recent advances in machine learning by deep neural networks are significant. But using these networks has been accompanied by a huge number of parameters for storage and computations that leads to an increase in the hardware cost and posing challenges. Therefore, compression approaches have been proposed to design efficient accelerators. One important approach for deep neural network compression is quantization that full-precision values are stored in low bit-width. In this way, in addition to memory saving, the operations will be replaced by simple ones with low cost. Many methods are suggested for DNNs Quantization in recent years, because of flexibility and influence in designing efficient hardware. Therefore, an integrated report is essential for better understanding, analysis, and comparison. In this paper, we provide a comprehensive survey. We describe the quantization concepts and categorize the methods from different perspectives. We discuss using the scale factor to match the quantization levels with the distribution of the full-precision values and describe the clustering-based methods. For the first time, we review the training of a quantized deep neural network and using Straight-Through Estimator comprehensively. Also, we describe the simplicity of operations in quantized deep convolutional neural networks and explain the sensitivity of the different layers in quantization. Finally, we discuss the evaluation of the quantization methods and compare the accuracy of previous methods with various bit-width for weights and activations on CIFAR-10 and the large-scale dataset, ImageNet.

Spiking Approximations of the MaxPooling Operation in Deep SNNs

Spiking Neural Networks (SNNs) are an emerging domain of biologically inspired neural networks that have shown promise for low-power AI. A number of methods exist for building deep SNNs, with Artificial Neural Network (ANN)-to-SNN conversion being highly successful. MaxPooling layers in Convolutional Neural Networks (CNNs) are an integral component to downsample the intermediate feature maps and introduce translational invariance, but the absence of their hardware-friendly spiking equivalents limits such CNNs' conversion to deep SNNs. In this paper, we present two hardware-friendly methods to implement Max-Pooling in deep SNNs, thus facilitating easy conversion of CNNs with MaxPooling layers to SNNs. In a first, we also execute SNNs with spiking-MaxPooling layers on Intel's Loihi neuromorphic hardware (with MNIST, FMNIST, & CIFAR10 dataset); thus, showing the feasibility of our approach.

A Learning Approach for Joint Design of Event-triggered Control and Power-Efficient Resource Allocation

In emerging Industrial Cyber-Physical Systems (ICPSs), the joint design of communication and control sub-systems is essential, as these sub-systems are interconnected. In this paper, we study the joint design problem of an event-triggered control and an energy-efficient resource allocation in a fifth generation (5G) wireless network. We formally state the problem as a multi-objective optimization one, aiming to minimize the number of updates on the actuators' input and the power consumption in the downlink transmission. To address the problem, we propose a model-free hierarchical reinforcement learning approach \textcolor{blue}{with uniformly ultimate boundedness stability guarantee} that learns four policies simultaneously. These policies contain an update time policy on the actuators' input, a control policy, and energy-efficient sub-carrier and power allocation policies. Our simulation results show that the proposed approach can properly control a simulated ICPS and significantly decrease the number of updates on the actuators' input as well as the downlink power consumption.

MIND: Maximum Mutual Information Based Neural Decoder

We are assisting at a growing interest in the development of learning architectures with application to digital communication systems. Herein, we consider the detection/decoding problem. We aim at developing an optimal neural architecture for such a task. The definition of the optimal criterion is a fundamental step. We propose to use the mutual information (MI) of the channel input-output signal pair. The computation of the MI is a formidable task, and for the majority of communication channels it is unknown. Therefore, the MI has to be learned. For such an objective, we propose a novel neural MI estimator based on a discriminative formulation. This leads to the derivation of the mutual information neural decoder (MIND). The developed neural architecture is capable not only to solve the decoding problem in unknown channels, but also to return an estimate of the average MI achieved with the coding scheme, as well as the decoding error probability. Several numerical results are reported and compared with maximum a-posteriori (MAP) and maximum likelihood (MaxL) decoding strategies.

GAN-Aimbots: Using Machine Learning for Cheating in First Person Shooters

Playing games with cheaters is not fun, and in a multi-billion-dollar video game industry with hundreds of millions of players, game developers aim to improve the security and, consequently, the user experience of their games by preventing cheating. Both traditional software-based methods and statistical systems have been successful in protecting against cheating, but recent advances in the automatic generation of content, such as images or speech, threaten the video game industry; they could be used to generate artificial gameplay indistinguishable from that of legitimate human players. To better understand this threat, we begin by reviewing the current state of multiplayer video game cheating, and then proceed to build a proof-of-concept method, GAN-Aimbot. By gathering data from various players in a first-person shooter game we show that the method improves players' performance while remaining hidden from automatic and manual protection mechanisms. By sharing this work we hope to raise awareness on this issue and encourage further research into protecting the gaming communities.

Generalization error bounds for DECONET: a deep unfolding network for analysis Compressive Sensing

In this paper, we propose a new deep unfolding neural network -- based on a state-of-the-art optimization algorithm -- for analysis Compressed Sensing. The proposed network called Decoding Network (DECONET) implements a decoder that reconstructs vectors from their incomplete, noisy measurements. Moreover, DECONET jointly learns a redundant analysis operator for sparsification, which is shared across the layers of DECONET. We study the generalization ability of DECONET. Towards that end, we first estimate the Rademacher complexity of the hypothesis class consisting of all the decoders that DECONET can implement. Then, we provide generalization error bounds, in terms of the aforementioned estimate. Finally, we present numerical experiments which confirm the validity of our theoretical results.

Fake News Quick Detection on Dynamic Heterogeneous Information Networks

The spread of fake news has caused great harm to society in recent years. So the quick detection of fake news has become an important task. Some current detection methods often model news articles and other related components as a static heterogeneous information network (HIN) and use expensive message-passing algorithms. However, in the real-world, quickly identifying fake news is of great significance and the network may vary over time in terms of dynamic nodes and edges. Therefore, in this paper, we propose a novel Dynamic Heterogeneous Graph Neural Network (DHGNN) for fake news quick detection. More specifically, we first implement BERT and fine-tuned BERT to get a semantic representation of the news article contents and author profiles and convert it into graph data. Then, we construct the heterogeneous news-author graph to reflect contextual information and relationships. Additionally, we adapt ideas from personalized PageRank propagation and dynamic propagation to heterogeneous networks in order to reduce the time complexity of back-propagating through many nodes during training. Experiments on three real-world fake news datasets show that DHGNN can outperform other GNN-based models in terms of both effectiveness and efficiency.

High Performance of Gradient Boosting in Binding Affinity Prediction

Prediction of protein-ligand (PL) binding affinity remains the key to drug discovery. Popular approaches in recent years involve graph neural networks (GNNs), which are used to learn the topology and geometry of PL complexes. However, GNNs are computationally heavy and have poor scalability to graph sizes. On the other hand, traditional machine learning (ML) approaches, such as gradient-boosted decision trees (GBDTs), are lightweight yet extremely efficient for tabular data. We propose to use PL interaction features along with PL graph-level features in GBDT. We show that this combination outperforms the existing solutions.

Cliff Diving: Exploring Reward Surfaces in Reinforcement Learning Environments

Visualizing optimization landscapes has led to many fundamental insights in numeric optimization, and novel improvements to optimization techniques. However, visualizations of the objective that reinforcement learning optimizes (the "reward surface") have only ever been generated for a small number of narrow contexts. This work presents reward surfaces and related visualizations of 27 of the most widely used reinforcement learning environments in Gym for the first time. We also explore reward surfaces in the policy gradient direction and show for the first time that many popular reinforcement learning environments have frequent "cliffs" (sudden large drops in expected return). We demonstrate that A2C often "dives off" these cliffs into low reward regions of the parameter space while PPO avoids them, confirming a popular intuition for PPO's improved performance over previous methods. We additionally introduce a highly extensible library that allows researchers to easily generate these visualizations in the future. Our findings provide new intuition to explain the successes and failures of modern RL methods, and our visualizations concretely characterize several failure modes of reinforcement learning agents in novel ways.

SaiNet: Stereo aware inpainting behind objects with generative networks

In this work, we present an end-to-end network for stereo-consistent image inpainting with the objective of inpainting large missing regions behind objects. The proposed model consists of an edge-guided UNet-like network using Partial Convolutions. We enforce multi-view stereo consistency by introducing a disparity loss. More importantly, we develop a training scheme where the model is learned from realistic stereo masks representing object occlusions, instead of the more common random masks. The technique is trained in a supervised way. Our evaluation shows competitive results compared to previous state-of-the-art techniques.

Integration of Text and Graph-based Features for Detecting Mental Health Disorders from Voice

With the availability of voice-enabled devices such as smart phones, mental health disorders could be detected and treated earlier, particularly post-pandemic. The current methods involve extracting features directly from audio signals. In this paper, two methods are used to enrich voice analysis for depression detection: graph transformation of voice signals, and natural language processing of the transcript based on representational learning, fused together to produce final class labels. The results of experiments with the DAIC-WOZ dataset suggest that integration of text-based voice classification and learning from low level and graph-based voice signal features can improve the detection of mental disorders like depression.

PrefixRL: Optimization of Parallel Prefix Circuits using Deep Reinforcement Learning

In this work, we present a reinforcement learning (RL) based approach to designing parallel prefix circuits such as adders or priority encoders that are fundamental to high-performance digital design. Unlike prior methods, our approach designs solutions tabula rasa purely through learning with synthesis in the loop. We design a grid-based state-action representation and an RL environment for constructing legal prefix circuits. Deep Convolutional RL agents trained on this environment produce prefix adder circuits that Pareto-dominate existing baselines with up to 16.0% and 30.2% lower area for the same delay in the 32b and 64b settings respectively. We observe that agents trained with open-source synthesis tools and cell library can design adder circuits that achieve lower area and delay than commercial tool adders in an industrial cell library.

Verifying Neural Networks Against Backdoor Attacks

Neural networks have achieved state-of-the-art performance in solving many problems, including many applications in safety/security-critical systems. Researchers also discovered multiple security issues associated with neural networks. One of them is backdoor attacks, i.e., a neural network may be embedded with a backdoor such that a target output is almost always generated in the presence of a trigger. Existing defense approaches mostly focus on detecting whether a neural network is 'backdoored' based on heuristics, e.g., activation patterns. To the best of our knowledge, the only line of work which certifies the absence of backdoor is based on randomized smoothing, which is known to significantly reduce neural network performance. In this work, we propose an approach to verify whether a given neural network is free of backdoor with a certain level of success rate. Our approach integrates statistical sampling as well as abstract interpretation. The experiment results show that our approach effectively verifies the absence of backdoor or generates backdoor triggers.

RASAT: Integrating Relational Structures into Pretrained Seq2Seq Model for Text-to-SQL

Relational structures such as schema linking and schema encoding have been validated as a key component to qualitatively translating natural language into SQL queries. However, introducing these structural relations comes with prices: they often result in a specialized model structure, which largely prohibits the use of large pretrained models in text-to-SQL. To address this problem, we propose RASAT: a Transformer seq2seq architecture augmented with relation-aware self-attention that could leverage a variety of relational structures while at the meantime being able to effectively inherit the pretrained parameters from the T5 model. Our model is able to incorporate almost all types of existing relations in the literature, and in addition, we propose to introduce co-reference relations for the multi-turn scenario. Experimental results on three widely used text-to-SQL datasets, covering both single-turn and multi-turn scenarios, have shown that RASAT could achieve competitive results in all three benchmarks, achieving state-of-the-art performance in execution accuracy (80.5\% EX on Spider, 53.1\% IEX on SParC, and 37.5\% IEX on CoSQL).

QHD: A brain-inspired hyperdimensional reinforcement learning algorithm

Reinforcement Learning (RL) has opened up new opportunities to solve a wide range of complex decision-making tasks. However, modern RL algorithms, e.g., Deep Q-Learning, are based on deep neural networks, putting high computational costs when running on edge devices. In this paper, we propose QHD, a Hyperdimensional Reinforcement Learning, that mimics brain properties toward robust and real-time learning. QHD relies on a lightweight brain-inspired model to learn an optimal policy in an unknown environment. We first develop a novel mathematical foundation and encoding module that maps state-action space into high-dimensional space. We accordingly develop a hyperdimensional regression model to approximate the Q-value function. The QHD-powered agent makes decisions by comparing Q-values of each possible action. We evaluate the effect of the different RL training batch sizes and local memory capacity on the QHD quality of learning. Our QHD is also capable of online learning with tiny local memory capacity, which can be as small as the training batch size. QHD provides real-time learning by further decreasing the memory capacity and the batch size. This makes QHD suitable for highly-efficient reinforcement learning in the edge environment, where it is crucial to support online and real-time learning. Our solution also supports a small experience replay batch size that provides 12.3 times speedup compared to DQN while ensuring minimal quality loss. Our evaluation shows QHD capability for real-time learning, providing 34.6 times speedup and significantly better quality of learning than state-of-the-art deep RL algorithms.

RiCS: A 2D Self-Occlusion Map for Harmonizing Volumetric Objects

There have been remarkable successes in computer vision with deep learning. While such breakthroughs show robust performance, there have still been many challenges in learning in-depth knowledge, like occlusion or predicting physical interactions. Although some recent works show the potential of 3D data in serving such context, it is unclear how we efficiently provide 3D input to the 2D models due to the misalignment in dimensionality between 2D and 3D. To leverage the successes of 2D models in predicting self-occlusions, we design Ray-marching in Camera Space (RiCS), a new method to represent the self-occlusions of foreground objects in 3D into a 2D self-occlusion map. We test the effectiveness of our representation on the human image harmonization task by predicting shading that is coherent with a given background image. Our experiments demonstrate that our representation map not only allows us to enhance the image quality but also to model temporally coherent complex shadow effects compared with the simulation-to-real and harmonization methods, both quantitatively and qualitatively. We further show that we can significantly improve the performance of human parts segmentation networks trained on existing synthetic datasets by enhancing the harmonization quality with our method.

Mask CycleGAN: Unpaired Multi-modal Domain Translation with Interpretable Latent Variable

We propose Mask CycleGAN, a novel architecture for unpaired image domain translation built based on CycleGAN, with an aim to address two issues: 1) unimodality in image translation and 2) lack of interpretability of latent variables. Our innovation in the technical approach is comprised of three key components: masking scheme, generator and objective. Experimental results demonstrate that this architecture is capable of bringing variations to generated images in a controllable manner and is reasonably robust to different masks.

No-regret learning for repeated non-cooperative games with lossy bandits

This paper considers no-regret learning for repeated continuous-kernel games with lossy bandit feedback. Since it is difficult to give the explicit model of the utility functions in dynamic environments, the players' action can only be learned with bandit feedback. Moreover, because of unreliable communication channels or privacy protection, the bandit feedback may be lost or dropped at random. Therefore, we study the asynchronous online learning strategy of the players to adaptively adjust the next actions for minimizing the long-term regret loss. The paper provides a novel no-regret learning algorithm, called Online Gradient Descent with lossy bandits (OGD-lb). We first give the regret analysis for concave games with differentiable and Lipschitz utilities. Then we show that the action profile converges to a Nash equilibrium with probability 1 when the game is also strictly monotone. We further provide the mean square convergence rate $\mathcal{O}\left(k^{-2\min\{\beta, 1/6\}}\right)$ when the game is $\beta-$ strongly monotone. In addition, we extend the algorithm to the case when the loss probability of the bandit feedback is unknown, and prove its almost sure convergence to Nash equilibrium for strictly monotone games. Finally, we take the resource management in fog computing as an application example, and carry out numerical experiments to empirically demonstrate the algorithm performance.

Improved Consistency Training for Semi-Supervised Sequence-to-Sequence ASR via Speech Chain Reconstruction and Self-Transcribing

Consistency regularization has recently been applied to semi-supervised sequence-to-sequence (S2S) automatic speech recognition (ASR). This principle encourages an ASR model to output similar predictions for the same input speech with different perturbations. The existing paradigm of semi-supervised S2S ASR utilizes SpecAugment as data augmentation and requires a static teacher model to produce pseudo transcripts for untranscribed speech. However, this paradigm fails to take full advantage of consistency regularization. First, the masking operations of SpecAugment may damage the linguistic contents of the speech, thus influencing the quality of pseudo labels. Second, S2S ASR requires both input speech and prefix tokens to make the next prediction. The static prefix tokens made by the offline teacher model cannot match dynamic pseudo labels during consistency training. In this work, we propose an improved consistency training paradigm of semi-supervised S2S ASR. We utilize speech chain reconstruction as the weak augmentation to generate high-quality pseudo labels. Moreover, we demonstrate that dynamic pseudo transcripts produced by the student ASR model benefit the consistency training. Experiments on LJSpeech and LibriSpeech corpora show that compared to supervised baselines, our improved paradigm achieves a 12.2% CER improvement in the single-speaker setting and 38.6% in the multi-speaker setting.

Bayesian Physics-Informed Extreme Learning Machine for Forward and Inverse PDE Problems with Noisy Data

Physics-informed extreme learning machine (PIELM) has recently received significant attention as a rapid version of physics-informed neural network (PINN) for solving partial differential equations (PDEs). The key characteristic is to fix the input layer weights with random values and use Moore-Penrose generalized inverse for the output layer weights. The framework is effective, but it easily suffers from overfitting noisy data and lacks uncertainty quantification for the solution under noise scenarios.To this end, we develop the Bayesian physics-informed extreme learning machine (BPIELM) to solve both forward and inverse linear PDE problems with noisy data in a unified framework. In our framework, a prior probability distribution is introduced in the output layer for extreme learning machine with physic laws and the Bayesian method is used to estimate the posterior of parameters. Besides, for inverse PDE problems, problem parameters considered as new output layer weights are unified in a framework with forward PDE problems. Finally, we demonstrate BPIELM considering both forward problems, including Poisson, advection, and diffusion equations, as well as inverse problems, where unknown problem parameters are estimated. The results show that, compared with PIELM, BPIELM quantifies uncertainty arising from noisy data and provides more accurate predictions. In addition, BPIELM is considerably cheaper than PINN in terms of the computational cost.

Unified Distributed Environment

We propose Unified Distributed Environment (UDE), an environment virtualization toolkit for reinforcement learning research. UDE is designed to integrate environments built on any simulation platform such as Gazebo, Unity, Unreal, and OpenAI Gym. Through environment virtualization, UDE enables offloading the environment for execution on a remote machine while still maintaining a unified interface. The UDE interface is designed to support multi-agent by default. With environment virtualization and its interface design, the agent policies can be trained in multiple machines for a multi-agent environment. Furthermore, UDE supports integration with existing major RL toolkits for researchers to leverage the benefits. This paper discusses the components of UDE and its design decisions.

Efficient Learning of Interpretable Classification Rules

Machine learning has become omnipresent with applications in various safety-critical domains such as medical, law, and transportation. In these domains, high-stake decisions provided by machine learning necessitate researchers to design interpretable models, where the prediction is understandable to a human. In interpretable machine learning, rule-based classifiers are particularly effective in representing the decision boundary through a set of rules comprising input features. The interpretability of rule-based classifiers is in general related to the size of the rules, where smaller rules are considered more interpretable. To learn such a classifier, the brute-force direct approach is to consider an optimization problem that tries to learn the smallest classification rule that has close to maximum accuracy. This optimization problem is computationally intractable due to its combinatorial nature and thus, the problem is not scalable in large datasets. To this end, in this paper we study the triangular relationship among the accuracy, interpretability, and scalability of learning rule-based classifiers. The contribution of this paper is an interpretable learning framework IMLI, that is based on maximum satisfiability (MaxSAT) for synthesizing classification rules expressible in proposition logic. Despite the progress of MaxSAT solving in the last decade, the straightforward MaxSAT-based solution cannot scale. Therefore, we incorporate an efficient incremental learning technique inside the MaxSAT formulation by integrating mini-batch learning and iterative rule-learning. In our experiments, IMLI achieves the best balance among prediction accuracy, interpretability, and scalability. As an application, we deploy IMLI in learning popular interpretable classifiers such as decision lists and decision sets.

Visual Exploration of Large-Scale Image Datasets for Machine Learning with Treemaps

In this paper, we present DendroMap, a novel approach to interactively exploring large-scale image datasets for machine learning. Machine learning practitioners often explore image datasets by generating a grid of images or projecting high-dimensional representations of images into 2-D using dimensionality reduction techniques (e.g., t-SNE). However, neither approach effectively scales to large datasets because images are ineffectively organized and interactions are insufficiently supported. To address these challenges, we develop DendroMap by adapting Treemaps, a well-known visualization technique. DendroMap effectively organizes images by extracting hierarchical cluster structures from high-dimensional representations of images. It enables users to make sense of the overall distributions of datasets and interactively zoom into specific areas of interests at multiple levels of abstraction. Our case studies with widely-used image datasets for deep learning demonstrate that users can discover insights about datasets and trained models by examining the diversity of images, identifying underperforming subgroups, and analyzing classification errors. We conducted a user study that evaluates the effectiveness of DendroMap in grouping and searching tasks by comparing it with a gridified version of t-SNE and found that participants preferred DendroMap over the compared method.

Toward a Geometrical Understanding of Self-supervised Contrastive Learning

Self-supervised learning (SSL) is currently one of the premier techniques to create data representations that are actionable for transfer learning in the absence of human annotations. Despite their success, the underlying geometry of these representations remains elusive, which obfuscates the quest for more robust, trustworthy, and interpretable models. In particular, mainstream SSL techniques rely on a specific deep neural network architecture with two cascaded neural networks: the encoder and the projector. When used for transfer learning, the projector is discarded since empirical results show that its representation generalizes more poorly than the encoder's. In this paper, we investigate this curious phenomenon and analyze how the strength of the data augmentation policies affects the data embedding. We discover a non-trivial relation between the encoder, the projector, and the data augmentation strength: with increasingly larger augmentation policies, the projector, rather than the encoder, is more strongly driven to become invariant to the augmentations. It does so by eliminating crucial information about the data by learning to project it into a low-dimensional space, a noisy estimate of the data manifold tangent plane in the encoder representation. This analysis is substantiated through a geometrical perspective with theoretical and empirical results.

A Tale of Two Flows: Cooperative Learning of Langevin Flow and Normalizing Flow Toward Energy-Based Model

This paper studies the cooperative learning of two generative flow models, in which the two models are iteratively updated based on the jointly synthesized examples. The first flow model is a normalizing flow that transforms an initial simple density to a target density by applying a sequence of invertible transformations. The second flow model is a Langevin flow that runs finite steps of gradient-based MCMC toward an energy-based model. We start from proposing a generative framework that trains an energy-based model with a normalizing flow as an amortized sampler to initialize the MCMC chains of the energy-based model. In each learning iteration, we generate synthesized examples by using a normalizing flow initialization followed by a short-run Langevin flow revision toward the current energy-based model. Then we treat the synthesized examples as fair samples from the energy-based model and update the model parameters with the maximum likelihood learning gradient, while the normalizing flow directly learns from the synthesized examples by maximizing the tractable likelihood. Under the short-run non-mixing MCMC scenario, the estimation of the energy-based model is shown to follow the perturbation of maximum likelihood, and the short-run Langevin flow and the normalizing flow form a two-flow generator that we call CoopFlow. We provide an understating of the CoopFlow algorithm by information geometry and show that it is a valid generator as it converges to a moment matching estimator. We demonstrate that the trained CoopFlow is capable of synthesizing realistic images, reconstructing images, and interpolating between images.

Exploring How Machine Learning Practitioners (Try To) Use Fairness Toolkits

Recent years have seen the development of many open-source ML fairness toolkits aimed at helping ML practitioners assess and address unfairness in their systems. However, there has been little research investigating how ML practitioners actually use these toolkits in practice. In this paper, we conducted the first in-depth empirical exploration of how industry practitioners (try to) work with existing fairness toolkits. In particular, we conducted think-aloud interviews to understand how participants learn about and use fairness toolkits, and explored the generality of our findings through an anonymous online survey. We identified several opportunities for fairness toolkits to better address practitioner needs and scaffold them in using toolkits effectively and responsibly. Based on these findings, we highlight implications for the design of future open-source fairness toolkits that can support practitioners in better contextualizing, communicating, and collaborating around ML fairness efforts.

Beyond General Purpose Machine Translation: The Need for Context-specific Empirical Research to Design for Appropriate User Trust

Machine Translation (MT) has the potential to help people overcome language barriers and is widely used in high-stakes scenarios, such as in hospitals. However, in order to use MT reliably and safely, users need to understand when to trust MT outputs and how to assess the quality of often imperfect translation results. In this paper, we discuss research directions to support users to calibrate trust in MT systems. We share findings from an empirical study in which we conducted semi-structured interviews with 20 clinicians to understand how they communicate with patients across language barriers, and if and how they use MT systems. Based on our findings, we advocate for empirical research on how MT systems are used in practice as an important first step to addressing the challenges in building appropriate trust between users and MT tools.

Representation learning with function call graph transformations for malware open set recognition

Open set recognition (OSR) problem has been a challenge in many machine learning (ML) applications, such as security. As new/unknown malware families occur regularly, it is difficult to exhaust samples that cover all the classes for the training process in ML systems. An advanced malware classification system should classify the known classes correctly while sensitive to the unknown class. In this paper, we introduce a self-supervised pre-training approach for the OSR problem in malware classification. We propose two transformations for the function call graph (FCG) based malware representations to facilitate the pretext task. Also, we present a statistical thresholding approach to find the optimal threshold for the unknown class. Moreover, the experiment results indicate that our proposed pre-training process can improve different performances of different downstream loss functions for the OSR problem.

Formal limitations of sample-wise information-theoretic generalization bounds

Some of the tightest information-theoretic generalization bounds depend on the average information between the learned hypothesis and a \emph{single} training example. However, these sample-wise bounds were derived only for \emph{expected} generalization gap. We show that even for expected \emph{squared} generalization gap no such sample-wise information-theoretic bounds exist. The same is true for PAC-Bayes and single-draw bounds. Remarkably, PAC-Bayes, single-draw and expected squared generalization gap bounds that depend on information in pairs of examples exist.

Neural-Fly Enables Rapid Learning for Agile Flight in Strong Winds

Executing safe and precise flight maneuvers in dynamic high-speed winds is important for the ongoing commoditization of uninhabited aerial vehicles (UAVs). However, because the relationship between various wind conditions and its effect on aircraft maneuverability is not well understood, it is challenging to design effective robot controllers using traditional control design methods. We present Neural-Fly, a learning-based approach that allows rapid online adaptation by incorporating pretrained representations through deep learning. Neural-Fly builds on two key observations that aerodynamics in different wind conditions share a common representation and that the wind-specific part lies in a low-dimensional space. To that end, Neural-Fly uses a proposed learning algorithm, domain adversarially invariant meta-learning (DAIML), to learn the shared representation, only using 12 minutes of flight data. With the learned representation as a basis, Neural-Fly then uses a composite adaptation law to update a set of linear coefficients for mixing the basis elements. When evaluated under challenging wind conditions generated with the Caltech Real Weather Wind Tunnel, with wind speeds up to 43.6 kilometers/hour (12.1 meters/second), Neural-Fly achieves precise flight control with substantially smaller tracking error than state-of-the-art nonlinear and adaptive controllers. In addition to strong empirical performance, the exponential stability of Neural-Fly results in robustness guarantees. Last, our control design extrapolates to unseen wind conditions, is shown to be effective for outdoor flights with only onboard sensors, and can transfer across drones with minimal performance degradation.

Multimodal Conversational AI: A Survey of Datasets and Approaches

As humans, we experience the world with all our senses or modalities (sound, sight, touch, smell, and taste). We use these modalities, particularly sight and touch, to convey and interpret specific meanings. Multimodal expressions are central to conversations; a rich set of modalities amplify and often compensate for each other. A multimodal conversational AI system answers questions, fulfills tasks, and emulates human conversations by understanding and expressing itself via multiple modalities. This paper motivates, defines, and mathematically formulates the multimodal conversational research objective. We provide a taxonomy of research required to solve the objective: multimodal representation, fusion, alignment, translation, and co-learning. We survey state-of-the-art datasets and approaches for each research area and highlight their limiting assumptions. Finally, we identify multimodal co-learning as a promising direction for multimodal conversational AI research.

Structural Dropout for Model Width Compression

Existing ML models are known to be highly over-parametrized, and use significantly more resources than required for a given task. Prior work has explored compressing models offline, such as by distilling knowledge from larger models into much smaller ones. This is effective for compression, but does not give an empirical method for measuring how much the model can be compressed, and requires additional training for each compressed model. We propose a method that requires only a single training session for the original model and a set of compressed models. The proposed approach is a "structural" dropout that prunes all elements in the hidden state above a randomly chosen index, forcing the model to learn an importance ordering over its features. After learning this ordering, at inference time unimportant features can be pruned while retaining most accuracy, reducing parameter size significantly. In this work, we focus on Structural Dropout for fully-connected layers, but the concept can be applied to any kind of layer with unordered features, such as convolutional or attention layers. Structural Dropout requires no additional pruning/retraining, but requires additional validation for each possible hidden sizes. At inference time, a non-expert can select a memory versus accuracy trade-off that best suits their needs, across a wide range of highly compressed versus more accurate models.

Perspectives on Incorporating Expert Feedback into Model Updates

Machine learning (ML) practitioners are increasingly tasked with developing models that are aligned with non-technical experts' values and goals. However, there has been insufficient consideration on how practitioners should translate domain expertise into ML updates. In this paper, we consider how to capture interactions between practitioners and experts systematically. We devise a taxonomy to match expert feedback types with practitioner updates. A practitioner may receive feedback from an expert at the observation- or domain-level, and convert this feedback into updates to the dataset, loss function, or parameter space. We review existing work from ML and human-computer interaction to describe this feedback-update taxonomy, and highlight the insufficient consideration given to incorporating feedback from non-technical experts. We end with a set of open questions that naturally arise from our proposed taxonomy and subsequent survey.

Developing a Production System for Purpose of Call Detection in Business Phone Conversations

For agents at a contact centre receiving calls, the most important piece of information is the reason for a given call. An agent cannot provide support on a call if they do not know why a customer is calling. In this paper we describe our implementation of a commercial system to detect Purpose of Call statements in English business call transcripts in real time. We present a detailed analysis of types of Purpose of Call statements and language patterns related to them, discuss an approach to collect rich training data by bootstrapping from a set of rules to a neural model, and describe a hybrid model which consists of a transformer-based classifier and a set of rules by leveraging insights from the analysis of call transcripts. The model achieved 88.6 F1 on average in various types of business calls when tested on real life data and has low inference time. We reflect on the challenges and design decisions when developing and deploying the system.

Universal Post-Training Backdoor Detection

A Backdoor attack (BA) is an important type of adversarial attack against deep neural network classifiers, wherein test samples from one or more source classes will be (mis)classified to the attacker's target class when a backdoor pattern (BP) is embedded. In this paper, we focus on the post-training backdoor defense scenario commonly considered in the literature, where the defender aims to detect whether a trained classifier was backdoor attacked, without any access to the training set. To the best of our knowledge, existing post-training backdoor defenses are all designed for BAs with presumed BP types, where each BP type has a specific embedding function. They may fail when the actual BP type used by the attacker (unknown to the defender) is different from the BP type assumed by the defender. In contrast, we propose a universal post-training defense that detects BAs with arbitrary types of BPs, without making any assumptions about the BP type. Our detector leverages the influence of the BA, independently of the BP type, on the landscape of the classifier's outputs prior to the softmax layer. For each class, a maximum margin statistic is estimated using a set of random vectors; detection inference is then performed by applying an unsupervised anomaly detector to these statistics. Thus, our detector is also an advance relative to most existing post-training methods by not needing any legitimate clean samples, and can efficiently detect BAs with arbitrary numbers of source classes. These advantages of our detector over several state-of-the-art methods are demonstrated on four datasets, for three different types of BPs, and for a variety of attack configurations. Finally, we propose a novel, general approach for BA mitigation once a detection is made.