title
stringlengths 7
246
| abstract
stringlengths 3
3.31k
|
---|---|
Persistency of Excitation for Robustness of Neural Networks | When an online learning algorithm is used to estimate the unknown parameters of a model, the signals interacting with the parameter estimates should not decay too quickly for the optimal values to be discovered correctly. This requirement is referred to as persistency of excitation, and it arises in various contexts, such as optimization with stochastic gradient methods, exploration for multi-armed bandits, and adaptive control of dynamical systems. While training a neural network, the iterative optimization algorithm involved also creates an online learning problem, and consequently, correct estimation of the optimal parameters requires persistent excitation of the network weights. In this work, we analyze the dynamics of the gradient descent algorithm while training a two-layer neural network with two different loss functions, the squared-error loss and the cross-entropy loss; and we obtain conditions to guarantee persistent excitation of the network weights. We then show that these conditions are difficult to satisfy when a multi-layer network is trained for a classification task, for the signals in the intermediate layers of the network become low-dimensional during training and fail to remain persistently exciting. To provide a remedy, we delve into the classical regularization terms used for linear models, reinterpret them as a means to ensure persistent excitation of the model parameters, and propose an algorithm for neural networks by building an analogy. The results in this work shed some light on why adversarial examples have become a challenging problem for neural networks, why merely augmenting training data sets will not be an effective approach to address them, and why there may not exist a data-independent regularization term for neural networks, which involve only the model parameters but not the training data. |
Rank-one matrix estimation: analytic time evolution of gradient descent dynamics | We consider a rank-one symmetric matrix corrupted by additive noise. The rank-one matrix is formed by an $n$-component unknown vector on the sphere of radius $\sqrt{n}$, and we consider the problem of estimating this vector from the corrupted matrix in the high dimensional limit of $n$ large, by gradient descent for a quadratic cost function on the sphere. Explicit formulas for the whole time evolution of the overlap between the estimator and unknown vector, as well as the cost, are rigorously derived. In the long time limit we recover the well known spectral phase transition, as a function of the signal-to-noise ratio. The explicit formulas also allow to point out interesting transient features of the time evolution. Our analysis technique is based on recent progress in random matrix theory and uses local versions of the semi-circle law. |
LOSSGRAD: automatic learning rate in gradient descent | In this paper, we propose a simple, fast and easy to implement algorithm LOSSGRAD (locally optimal step-size in gradient descent), which automatically modifies the step-size in gradient descent during neural networks training. Given a function $f$, a point $x$, and the gradient $\nabla_x f$ of $f$, we aim to find the step-size $h$ which is (locally) optimal, i.e. satisfies: $$ h=arg\,min_{t \geq 0} f(x-t \nabla_x f). $$ Making use of quadratic approximation, we show that the algorithm satisfies the above assumption. We experimentally show that our method is insensitive to the choice of initial learning rate while achieving results comparable to other methods. |
Let the Model Decide its Curriculum for Multitask Learning | Curriculum learning strategies in prior multi-task learning approaches arrange datasets in a difficulty hierarchy either based on human perception or by exhaustively searching the optimal arrangement. However, human perception of difficulty may not always correlate well with machine interpretation leading to poor performance and exhaustive search is computationally expensive. Addressing these concerns, we propose two classes of techniques to arrange training instances into a learning curriculum based on difficulty scores computed via model-based approaches. The two classes i.e Dataset-level and Instance-level differ in granularity of arrangement. Through comprehensive experiments with 12 datasets, we show that instance-level and dataset-level techniques result in strong representations as they lead to an average performance improvement of 4.17% and 3.15% over their respective baselines. Furthermore, we find that most of this improvement comes from correctly answering the difficult instances, implying a greater efficacy of our techniques on difficult tasks. |
Explainable Deep Relational Networks for Predicting Compound-Protein Affinities and Contacts | Predicting compound-protein affinity is critical for accelerating drug discovery. Recent progress made by machine learning focuses on accuracy but leaves much to be desired for interpretability. Through molecular contacts underlying affinities, our large-scale interpretability assessment finds commonly-used attention mechanisms inadequate. We thus formulate a hierarchical multi-objective learning problem whose predicted contacts form the basis for predicted affinities. We further design a physics-inspired deep relational network, DeepRelations, with intrinsically explainable architecture. Specifically, various atomic-level contacts or "relations" lead to molecular-level affinity prediction. And the embedded attentions are regularized with predicted structural contexts and supervised with partially available training contacts. DeepRelations shows superior interpretability to the state-of-the-art: without compromising affinity prediction, it boosts the AUPRC of contact prediction 9.5, 16.9, 19.3 and 5.7-fold for the test, compound-unique, protein-unique, and both-unique sets, respectively. Our study represents the first dedicated model development and systematic model assessment for interpretable machine learning of compound-protein affinity. |
Symphony Generation with Permutation Invariant Language Model | In this work, we present a symbolic symphony music generation solution, SymphonyNet, based on a permutation invariant language model. To bridge the gap between text generation and symphony generation task, we propose a novel Multi-track Multi-instrument Repeatable (MMR) representation with particular 3-D positional embedding and a modified Byte Pair Encoding algorithm (Music BPE) for music tokens. A novel linear transformer decoder architecture is introduced as a backbone for modeling extra-long sequences of symphony tokens. Meanwhile, we train the decoder to learn automatic orchestration as a joint task by masking instrument information from the input. We also introduce a large-scale symbolic symphony dataset for the advance of symphony generation research. Our empirical results show that our proposed approach can generate coherent, novel, complex and harmonious symphony compared to human composition, which is the pioneer solution for multi-track multi-instrument symbolic music generation. |
Metrics for more than two points at once | The conventional definition of a topological metric over a space specifies properties that must be obeyed by any measure of "how separated" two points in that space are. Here it is shown how to extend that definition, and in particular the triangle inequality, to concern arbitrary numbers of points. Such a measure of how separated the points within a collection are can be bootstrapped, to measure "how separated" from each other are two (or more) collections. The measure presented here also allows fractional membership of an element in a collection. This means it directly concerns measures of ``how spread out" a probability distribution over a space is. When such a measure is bootstrapped to compare two collections, it allows us to measure how separated two probability distributions are, or more generally, how separated a distribution of distributions is. |
GAP-Gen: Guided Automatic Python Code Generation | Automatic code generation from natural language descriptions can be highly beneficial during the process of software development. In this work, we propose GAP-Gen, an automatic code generation method guided by Python syntactic constraints and semantic constraints. We first introduce Python syntactic constraints in the form of Syntax-Flow, which is a simplified version of Abstract Syntax Tree (AST) reducing the size and high complexity of Abstract Syntax Tree but maintaining the crucial syn-tactic information of Python code. In addition to Syntax-Flow, we introduce Variable-Flow which abstracts variable and function names consistently throughout the code. In our work, rather than pre-training, we focus on modifying the fine-tuning process which reduces computational requirements but retains high generation performance on automatic Python code generation task. GAP-Gen fine-tunes the transformer-based language models T5 and CodeT5 using the Code-to-Docstring datasets CodeSearchNet, CodeSearchNet AdvTest, and Code-Docstring-Corpus from EdinburghNLP. Our experiments show that GAP-Gen achieves better results on automatic Python code generation task than previous works |
STAD: Spatio-Temporal Adjustment of Traffic-Oblivious Travel-Time Estimation | Travel time estimation is an important component in modern transportation applications. The state of the art techniques for travel time estimation use GPS traces to learn the weights of a road network, often modeled as a directed graph, then apply Dijkstra-like algorithms to find shortest paths. Travel time is then computed as the sum of edge weights on the returned path. In order to enable time-dependency, existing systems compute multiple weighted graphs corresponding to different time windows. These graphs are often optimized offline before they are deployed into production routing engines, causing a serious engineering overhead. In this paper, we present STAD, a system that adjusts - on the fly - travel time estimates for any trip request expressed in the form of origin, destination, and departure time. STAD uses machine learning and sparse trips data to learn the imperfections of any basic routing engine, before it turns it into a full-fledged time-dependent system capable of adjusting travel times to real traffic conditions in a city. STAD leverages the spatio-temporal properties of traffic by combining spatial features such as departing and destination geographic zones with temporal features such as departing time and day to significantly improve the travel time estimates of the basic routing engine. Experiments on real trip datasets from Doha, New York City, and Porto show a reduction in median absolute errors of 14% in the first two cities and 29% in the latter. We also show that STAD performs better than different commercial and research baselines in all three cities. |
Learning with Bounded Instance- and Label-dependent Label Noise | Instance- and Label-dependent label Noise (ILN) widely exists in real-world datasets but has been rarely studied. In this paper, we focus on Bounded Instance- and Label-dependent label Noise (BILN), a particular case of ILN where the label noise rates -- the probabilities that the true labels of examples flip into the corrupted ones -- have upper bound less than $1$. Specifically, we introduce the concept of distilled examples, i.e. examples whose labels are identical with the labels assigned for them by the Bayes optimal classifier, and prove that under certain conditions classifiers learnt on distilled examples will converge to the Bayes optimal classifier. Inspired by the idea of learning with distilled examples, we then propose a learning algorithm with theoretical guarantees for its robustness to BILN. At last, empirical evaluations on both synthetic and real-world datasets show effectiveness of our algorithm in learning with BILN. |
Natural Backdoor Attack on Text Data | Recently, advanced NLP models have seen a surge in the usage of various applications. This raises the security threats of the released models. In addition to the clean models' unintentional weaknesses, {\em i.e.,} adversarial attacks, the poisoned models with malicious intentions are much more dangerous in real life. However, most existing works currently focus on the adversarial attacks on NLP models instead of positioning attacks, also named \textit{backdoor attacks}. In this paper, we first propose the \textit{natural backdoor attacks} on NLP models. Moreover, we exploit the various attack strategies to generate trigger on text data and investigate different types of triggers based on modification scope, human recognition, and special cases. Last, we evaluate the backdoor attacks, and the results show the excellent performance of with 100\% backdoor attacks success rate and sacrificing of 0.83\% on the text classification task. |
Multi-agent Reinforcement Learning Accelerated MCMC on Multiscale Inversion Problem | In this work, we propose a multi-agent actor-critic reinforcement learning (RL) algorithm to accelerate the multi-level Monte Carlo Markov Chain (MCMC) sampling algorithms. The policies (actors) of the agents are used to generate the proposal in the MCMC steps; and the critic, which is centralized, is in charge of estimating the long term reward. We verify our proposed algorithm by solving an inverse problem with multiple scales. There are several difficulties in the implementation of this problem by using traditional MCMC sampling. Firstly, the computation of the posterior distribution involves evaluating the forward solver, which is very time consuming for a problem with heterogeneous. We hence propose to use the multi-level algorithm. More precisely, we use the generalized multiscale finite element method (GMsFEM) as the forward solver in evaluating a posterior distribution in the multi-level rejection procedure. Secondly, it is hard to find a function which can generate samplings which are meaningful. To solve this issue, we learn an RL policy as the proposal generator. Our experiments show that the proposed method significantly improves the sampling process |
Machine learning approximation algorithms for high-dimensional fully nonlinear partial differential equations and second-order backward stochastic differential equations | High-dimensional partial differential equations (PDE) appear in a number of models from the financial industry, such as in derivative pricing models, credit valuation adjustment (CVA) models, or portfolio optimization models. The PDEs in such applications are high-dimensional as the dimension corresponds to the number of financial assets in a portfolio. Moreover, such PDEs are often fully nonlinear due to the need to incorporate certain nonlinear phenomena in the model such as default risks, transaction costs, volatility uncertainty (Knightian uncertainty), or trading constraints in the model. Such high-dimensional fully nonlinear PDEs are exceedingly difficult to solve as the computational effort for standard approximation methods grows exponentially with the dimension. In this work we propose a new method for solving high-dimensional fully nonlinear second-order PDEs. Our method can in particular be used to sample from high-dimensional nonlinear expectations. The method is based on (i) a connection between fully nonlinear second-order PDEs and second-order backward stochastic differential equations (2BSDEs), (ii) a merged formulation of the PDE and the 2BSDE problem, (iii) a temporal forward discretization of the 2BSDE and a spatial approximation via deep neural nets, and (iv) a stochastic gradient descent-type optimization procedure. Numerical results obtained using ${\rm T{\small ENSOR}F{\small LOW}}$ in ${\rm P{\small YTHON}}$ illustrate the efficiency and the accuracy of the method in the cases of a $100$-dimensional Black-Scholes-Barenblatt equation, a $100$-dimensional Hamilton-Jacobi-Bellman equation, and a nonlinear expectation of a $ 100 $-dimensional $ G $-Brownian motion. |
Iterative Reweighted Minimization Methods for $l_p$ Regularized Unconstrained Nonlinear Programming | In this paper we study general $l_p$ regularized unconstrained minimization problems. In particular, we derive lower bounds for nonzero entries of first- and second-order stationary points, and hence also of local minimizers of the $l_p$ minimization problems. We extend some existing iterative reweighted $l_1$ (IRL1) and $l_2$ (IRL2) minimization methods to solve these problems and proposed new variants for them in which each subproblem has a closed form solution. Also, we provide a unified convergence analysis for these methods. In addition, we propose a novel Lipschitz continuous $\epsilon$-approximation to $\|x\|^p_p$. Using this result, we develop new IRL1 methods for the $l_p$ minimization problems and showed that any accumulation point of the sequence generated by these methods is a first-order stationary point, provided that the approximation parameter $\epsilon$ is below a computable threshold value. This is a remarkable result since all existing iterative reweighted minimization methods require that $\epsilon$ be dynamically updated and approach zero. Our computational results demonstrate that the new IRL1 method is generally more stable than the existing IRL1 methods [21,18] in terms of objective function value and CPU time. |
Analytic Insights into Structure and Rank of Neural Network Hessian Maps | The Hessian of a neural network captures parameter interactions through second-order derivatives of the loss. It is a fundamental object of study, closely tied to various problems in deep learning, including model design, optimization, and generalization. Most prior work has been empirical, typically focusing on low-rank approximations and heuristics that are blind to the network structure. In contrast, we develop theoretical tools to analyze the range of the Hessian map, providing us with a precise understanding of its rank deficiency as well as the structural reasons behind it. This yields exact formulas and tight upper bounds for the Hessian rank of deep linear networks, allowing for an elegant interpretation in terms of rank deficiency. Moreover, we demonstrate that our bounds remain faithful as an estimate of the numerical Hessian rank, for a larger class of models such as rectified and hyperbolic tangent networks. Further, we also investigate the implications of model architecture (e.g.~width, depth, bias) on the rank deficiency. Overall, our work provides novel insights into the source and extent of redundancy in overparameterized networks. |
Unsupervised Representation Learning by InvariancePropagation | Unsupervised learning methods based on contrastive learning have drawn increasing attention and achieved promising results. Most of them aim to learn representations invariant to instance-level variations, which are provided by different views of the same instance. In this paper, we propose Invariance Propagation to focus on learning representations invariant to category-level variations, which are provided by different instances from the same category. Our method recursively discovers semantically consistent samples residing in the same high-density regions in representation space. We demonstrate a hard sampling strategy to concentrate on maximizing the agreement between the anchor sample and its hard positive samples, which provide more intra-class variations to help capture more abstract invariance. As a result, with a ResNet-50 as the backbone, our method achieves 71.3% top-1 accuracy on ImageNet linear classification and 78.2% top-5 accuracy fine-tuning on only 1% labels, surpassing previous results. We also achieve state-of-the-art performance on other downstream tasks, including linear classification on Places205 and Pascal VOC, and transfer learning on small scale datasets. |
Scalable MCMC for Mixed Membership Stochastic Blockmodels | We propose a stochastic gradient Markov chain Monte Carlo (SG-MCMC) algorithm for scalable inference in mixed-membership stochastic blockmodels (MMSB). Our algorithm is based on the stochastic gradient Riemannian Langevin sampler and achieves both faster speed and higher accuracy at every iteration than the current state-of-the-art algorithm based on stochastic variational inference. In addition we develop an approximation that can handle models that entertain a very large number of communities. The experimental results show that SG-MCMC strictly dominates competing algorithms in all cases. |
Evaluation of Spectral Learning for the Identification of Hidden Markov Models | Hidden Markov models have successfully been applied as models of discrete time series in many fields. Often, when applied in practice, the parameters of these models have to be estimated. The currently predominating identification methods, such as maximum-likelihood estimation and especially expectation-maximization, are iterative and prone to have problems with local minima. A non-iterative method employing a spectral subspace-like approach has recently been proposed in the machine learning literature. This paper evaluates the performance of this algorithm, and compares it to the performance of the expectation-maximization algorithm, on a number of numerical examples. We find that the performance is mixed; it successfully identifies some systems with relatively few available observations, but fails completely for some systems even when a large amount of observations is available. An open question is how this discrepancy can be explained. We provide some indications that it could be related to how well-conditioned some system parameters are. |
Deep Predictive Coding Networks for Video Prediction and Unsupervised Learning | While great strides have been made in using deep learning algorithms to solve supervised learning tasks, the problem of unsupervised learning - leveraging unlabeled examples to learn about the structure of a domain - remains a difficult unsolved challenge. Here, we explore prediction of future frames in a video sequence as an unsupervised learning rule for learning about the structure of the visual world. We describe a predictive neural network ("PredNet") architecture that is inspired by the concept of "predictive coding" from the neuroscience literature. These networks learn to predict future frames in a video sequence, with each layer in the network making local predictions and only forwarding deviations from those predictions to subsequent network layers. We show that these networks are able to robustly learn to predict the movement of synthetic (rendered) objects, and that in doing so, the networks learn internal representations that are useful for decoding latent object parameters (e.g. pose) that support object recognition with fewer training views. We also show that these networks can scale to complex natural image streams (car-mounted camera videos), capturing key aspects of both egocentric movement and the movement of objects in the visual scene, and the representation learned in this setting is useful for estimating the steering angle. Altogether, these results suggest that prediction represents a powerful framework for unsupervised learning, allowing for implicit learning of object and scene structure. |
Vehicle Trajectory Prediction by Transfer Learning of Semi-Supervised Models | In this work we show that semi-supervised models for vehicle trajectory prediction significantly improve performance over supervised models on state-of-the-art real-world benchmarks. Moving from supervised to semi-supervised models allows scaling-up by using unlabeled data, increasing the number of images in pre-training from Millions to a Billion. We perform ablation studies comparing transfer learning of semi-supervised and supervised models while keeping all other factors equal. Within semi-supervised models we compare contrastive learning with teacher-student methods as well as networks predicting a small number of trajectories with networks predicting probabilities over a large trajectory set. Our results using both low-level and mid-level representations of the driving environment demonstrate the applicability of semi-supervised methods for real-world vehicle trajectory prediction. |
Learning icons appearance similarity | Selecting an optimal set of icons is a crucial step in the pipeline of visual design to structure and navigate through content. However, designing the icons sets is usually a difficult task for which expert knowledge is required. In this work, to ease the process of icon set selection to the users, we propose a similarity metric which captures the properties of style and visual identity. We train a Siamese Neural Network with an online dataset of icons organized in visually coherent collections that are used to adaptively sample training data and optimize the training process. As the dataset contains noise, we further collect human-rated information on the perception of icon's similarity which will be used for evaluating and testing the proposed model. We present several results and applications based on searches, kernel visualizations and optimized set proposals that can be helpful for designers and non-expert users while exploring large collections of icons. |
Cross-lingual Offensive Language Identification for Low Resource Languages: The Case of Marathi | The widespread presence of offensive language on social media motivated the development of systems capable of recognizing such content automatically. Apart from a few notable exceptions, most research on automatic offensive language identification has dealt with English. To address this shortcoming, we introduce MOLD, the Marathi Offensive Language Dataset. MOLD is the first dataset of its kind compiled for Marathi, thus opening a new domain for research in low-resource Indo-Aryan languages. We present results from several machine learning experiments on this dataset, including zero-short and other transfer learning experiments on state-of-the-art cross-lingual transformers from existing data in Bengali, English, and Hindi. |
An Interpretable Neural Network for Parameter Inference | Adoption of deep neural networks in fields such as economics or finance has been constrained by the lack of interpretability of model outcomes. This paper proposes a generative neural network architecture - the parameter encoder neural network (PENN) - capable of estimating local posterior distributions for the parameters of a regression model. The parameters fully explain predictions in terms of the inputs and permit visualization, interpretation and inference in the presence of complex heterogeneous effects and feature dependencies. The use of Bayesian inference techniques offers an intuitive mechanism to regularize local parameter estimates towards a stable solution, and to reduce noise-fitting in settings of limited data availability. The proposed neural network is particularly well-suited to applications in economics and finance, where parameter inference plays an important role. An application to an asset pricing problem demonstrates how the PENN can be used to explore nonlinear risk dynamics in financial markets, and to compare empirical nonlinear effects to behavior posited by financial theory. |
Transfer Learning in Conversational Analysis through Reusing Preprocessing Data as Supervisors | Conversational analysis systems are trained using noisy human labels and often require heavy preprocessing during multi-modal feature extraction. Using noisy labels in single-task learning increases the risk of over-fitting. Auxiliary tasks could improve the performance of the primary task learning during the same training -- this approach sits in the intersection of transfer learning and multi-task learning (MTL). In this paper, we explore how the preprocessed data used for feature engineering can be re-used as auxiliary tasks, thereby promoting the productive use of data. Our main contributions are: (1) the identification of sixteen beneficially auxiliary tasks, (2) studying the method of distributing learning capacity between the primary and auxiliary tasks, and (3) studying the relative supervision hierarchy between the primary and auxiliary tasks. Extensive experiments on IEMOCAP and SEMAINE data validate the improvements over single-task approaches, and suggest that it may generalize across multiple primary tasks. |
Quantal synaptic dilution enhances sparse encoding and dropout regularisation in deep networks | Dropout is a technique that silences the activity of units stochastically while training deep networks to reduce overfitting. Here we introduce Quantal Synaptic Dilution (QSD), a biologically plausible model of dropout regularisation based on the quantal properties of neuronal synapses, that incorporates heterogeneities in response magnitudes and release probabilities for vesicular quanta. QSD outperforms standard dropout in ReLU multilayer perceptrons, with enhanced sparse encoding at test time when dropout masks are replaced with identity functions, without shifts in trainable weight or bias distributions. For convolutional networks, the method also improves generalisation in computer vision tasks with and without inclusion of additional forms of regularisation. QSD also outperforms standard dropout in recurrent networks for language modelling and sentiment analysis. An advantage of QSD over many variations of dropout is that it can be implemented generally in all conventional deep networks where standard dropout is applicable. |
DeepFlash: Turning a Flash Selfie into a Studio Portrait | We present a method for turning a flash selfie taken with a smartphone into a photograph as if it was taken in a studio setting with uniform lighting. Our method uses a convolutional neural network trained on a set of pairs of photographs acquired in an ad-hoc acquisition campaign. Each pair consists of one photograph of a subject's face taken with the camera flash enabled and another one of the same subject in the same pose illuminated using a photographic studio-lighting setup. We show how our method can amend defects introduced by a close-up camera flash, such as specular highlights, shadows, skin shine, and flattened images. |
Data-Free Adversarial Distillation | Knowledge Distillation (KD) has made remarkable progress in the last few years and become a popular paradigm for model compression and knowledge transfer. However, almost all existing KD algorithms are data-driven, i.e., relying on a large amount of original training data or alternative data, which is usually unavailable in real-world scenarios. In this paper, we devote ourselves to this challenging problem and propose a novel adversarial distillation mechanism to craft a compact student model without any real-world data. We introduce a model discrepancy to quantificationally measure the difference between student and teacher models and construct an optimizable upper bound. In our work, the student and the teacher jointly act the role of the discriminator to reduce this discrepancy, when a generator adversarially produces some "hard samples" to enlarge it. Extensive experiments demonstrate that the proposed data-free method yields comparable performance to existing data-driven methods. More strikingly, our approach can be directly extended to semantic segmentation, which is more complicated than classification, and our approach achieves state-of-the-art results. Code and pretrained models are available at https://github.com/VainF/Data-Free-Adversarial-Distillation. |
Towards a Solution to Bongard Problems: A Causal Approach | To date, Bongard Problems (BP) remain one of the few fortresses of AI history yet to be raided by the powerful models of the current era. We present a systematic analysis using modern techniques from the intersection of causality and AI/ML in a humble effort of reviving research around BPs. Specifically, we first compile the BPs into a Markov decision process, then secondly pose causal assumptions on the data generating process arguing for their applicability to BPs, and finally apply reinforcement learning techniques for solving the BPs subject to the causal assumptions. |
Vision-based Control of a Quadrotor in User Proximity: Mediated vs End-to-End Learning Approaches | We consider the task of controlling a quadrotor to hover in front of a freely moving user, using input data from an onboard camera. On this specific task we compare two widespread learning paradigms: a mediated approach, which learns an high-level state from the input and then uses it for deriving control signals; and an end-to-end approach, which skips high-level state estimation altogether. We show that despite their fundamental difference, both approaches yield equivalent performance on this task. We finally qualitatively analyze the behavior of a quadrotor implementing such approaches. |
Multiple-Instance Learning: Radon-Nikodym Approach to Distribution Regression Problem | For distribution regression problem, where a bag of $x$--observations is mapped to a single $y$ value, a one--step solution is proposed. The problem of random distribution to random value is transformed to random vector to random value by taking distribution moments of $x$ observations in a bag as random vector. Then Radon--Nikodym or least squares theory can be applied, what give $y(x)$ estimator. The probability distribution of $y$ is also obtained, what requires solving generalized eigenvalues problem, matrix spectrum (not depending on $x$) give possible $y$ outcomes and depending on $x$ probabilities of outcomes can be obtained by projecting the distribution with fixed $x$ value (delta--function) to corresponding eigenvector. A library providing numerically stable polynomial basis for these calculations is available, what make the proposed approach practical. |
Potential-Function Proofs for First-Order Methods | This note discusses proofs for convergence of first-order methods based on simple potential-function arguments. We cover methods like gradient descent (for both smooth and non-smooth settings), mirror descent, and some accelerated variants. |
Sensor Fusion using Backward Shortcut Connections for Sleep Apnea Detection in Multi-Modal Data | Sleep apnea is a common respiratory disorder characterized by breathing pauses during the night. Consequences of untreated sleep apnea can be severe. Still, many people remain undiagnosed due to shortages of hospital beds and trained sleep technicians. To assist in the diagnosis process, automated detection methods are being developed. Recent works have demonstrated that deep learning models can extract useful information from raw respiratory data and that such models can be used as a robust sleep apnea detector. However, trained sleep technicians take into account multiple sensor signals when annotating sleep recordings instead of relying on a single respiratory estimate. To improve the predictive performance and reliability of the models, early and late sensor fusion methods are explored in this work. In addition, a novel late sensor fusion method is proposed which uses backward shortcut connections to improve the learning of the first stages of the models. The performance of these fusion methods is analyzed using CNN as well as LSTM deep learning base-models. The results demonstrate a significant and consistent improvement in predictive performance over the single sensor methods and over the other explored sensor fusion methods, by using the proposed sensor fusion method with backward shortcut connections. |
Approximate Learning in Complex Dynamic Bayesian Networks | In this paper we extend the work of Smith and Papamichail (1999) and present fast approximate Bayesian algorithms for learning in complex scenarios where at any time frame, the relationships between explanatory state space variables can be described by a Bayesian network that evolve dynamically over time and the observations taken are not necessarily Gaussian. It uses recent developments in approximate Bayesian forecasting methods in combination with more familiar Gaussian propagation algorithms on junction trees. The procedure for learning state parameters from data is given explicitly for common sampling distributions and the methodology is illustrated through a real application. The efficiency of the dynamic approximation is explored by using the Hellinger divergence measure and theoretical bounds for the efficacy of such a procedure are discussed. |
Deep Algorithms: designs for networks | A new design methodology for neural networks that is guided by traditional algorithm design is presented. To prove our point, we present two heuristics and demonstrate an algorithmic technique for incorporating additional weights in their signal-flow graphs. We show that with training the performance of these networks can not only exceed the performance of the initial network, but can match the performance of more-traditional neural network architectures. A key feature of our approach is that these networks are initialized with parameters that provide a known performance threshold for the architecture on a given task. |
Deep Gaussian Processes with Convolutional Kernels | Deep Gaussian processes (DGPs) provide a Bayesian non-parametric alternative to standard parametric deep learning models. A DGP is formed by stacking multiple GPs resulting in a well-regularized composition of functions. The Bayesian framework that equips the model with attractive properties, such as implicit capacity control and predictive uncertainty, makes it at the same time challenging to combine with a convolutional structure. This has hindered the application of DGPs in computer vision tasks, an area where deep parametric models (i.e. CNNs) have made breakthroughs. Standard kernels used in DGPs such as radial basis functions (RBFs) are insufficient for handling pixel variability in raw images. In this paper, we build on the recent convolutional GP to develop Convolutional DGP (CDGP) models which effectively capture image level features through the use of convolution kernels, therefore opening up the way for applying DGPs to computer vision tasks. Our model learns local spatial influence and outperforms strong GP based baselines on multi-class image classification. We also consider various constructions of convolution kernel over the image patches, analyze the computational trade-offs and provide an efficient framework for convolutional DGP models. The experimental results on image data such as MNIST, rectangles-image, CIFAR10 and Caltech101 demonstrate the effectiveness of the proposed approaches. |
A Probabilistic framework for Quantum Clustering | Quantum Clustering is a powerful method to detect clusters in data with mixed density. However, it is very sensitive to a length parameter that is inherent to the Schr\"odinger equation. In addition, linking data points into clusters requires local estimates of covariance that are also controlled by length parameters. This raises the question of how to adjust the control parameters of the Schr\"odinger equation for optimal clustering. We propose a probabilistic framework that provides an objective function for the goodness-of-fit to the data, enabling the control parameters to be optimised within a Bayesian framework. This naturally yields probabilities of cluster membership and data partitions with specific numbers of clusters. The proposed framework is tested on real and synthetic data sets, assessing its validity by measuring concordance with known data structure by means of the Jaccard score (JS). This work also proposes an objective way to measure performance in unsupervised learning that correlates very well with JS. |
Density Matching Reward Learning | In this paper, we focus on the problem of inferring the underlying reward function of an expert given demonstrations, which is often referred to as inverse reinforcement learning (IRL). In particular, we propose a model-free density-based IRL algorithm, named density matching reward learning (DMRL), which does not require model dynamics. The performance of DMRL is analyzed theoretically and the sample complexity is derived. Furthermore, the proposed DMRL is extended to handle nonlinear IRL problems by assuming that the reward function is in the reproducing kernel Hilbert space (RKHS) and kernel DMRL (KDMRL) is proposed. The parameters for KDMRL can be computed analytically, which greatly reduces the computation time. The performance of KDMRL is extensively evaluated in two sets of experiments: grid world and track driving experiments. In grid world experiments, the proposed KDMRL method is compared with both model-based and model-free IRL methods and shows superior performance on a nonlinear reward setting and competitive performance on a linear reward setting in terms of expected value differences. Then we move on to more realistic experiments of learning different driving styles for autonomous navigation in complex and dynamic tracks using KDMRL and receding horizon control. |
Generative Chemical Transformer: Neural Machine Learning of Molecular Geometric Structures from Chemical Language via Attention | Discovering new materials better suited to specific purposes is an important issue in improving the quality of human life. Here, a neural network that creates molecules that meet some desired conditions based on a deep understanding of chemical language is proposed (Generative Chemical Transformer, GCT). The attention mechanism in GCT allows a deeper understanding of molecular structures beyond the limitations of chemical language itself which cause semantic discontinuity by paying attention to characters sparsely. It is investigated that the significance of language models for inverse molecular design problems by quantitatively evaluating the quality of the generated molecules. GCT generates highly realistic chemical strings that satisfy both chemical and linguistic grammar rules. Molecules parsed from generated strings simultaneously satisfy the multiple target properties and vary for a single condition set. These advances will contribute to improving the quality of human life by accelerating the process of desired material discovery. |
iGOS++: Integrated Gradient Optimized Saliency by Bilateral Perturbations | The black-box nature of the deep networks makes the explanation for "why" they make certain predictions extremely challenging. Saliency maps are one of the most widely-used local explanation tools to alleviate this problem. One of the primary approaches for generating saliency maps is by optimizing a mask over the input dimensions so that the output of the network is influenced the most by the masking. However, prior work only studies such influence by removing evidence from the input. In this paper, we present iGOS++, a framework to generate saliency maps that are optimized for altering the output of the black-box system by either removing or preserving only a small fraction of the input. Additionally, we propose to add a bilateral total variation term to the optimization that improves the continuity of the saliency map especially under high resolution and with thin object parts. The evaluation results from comparing iGOS++ against state-of-the-art saliency map methods show significant improvement in locating salient regions that are directly interpretable by humans. We utilized iGOS++ in the task of classifying COVID-19 cases from x-ray images and discovered that sometimes the CNN network is overfitted to the characters printed on the x-ray images when performing classification. Fixing this issue by data cleansing significantly improved the precision and recall of the classifier. |
SDCA without Duality, Regularization, and Individual Convexity | Stochastic Dual Coordinate Ascent is a popular method for solving regularized loss minimization for the case of convex losses. We describe variants of SDCA that do not require explicit regularization and do not rely on duality. We prove linear convergence rates even if individual loss functions are non-convex, as long as the expected loss is strongly convex. |
When is there a Representer Theorem? Nondifferentiable Regularisers and Banach spaces | We consider a general regularised interpolation problem for learning a parameter vector from data. The well known representer theorem says that under certain conditions on the regulariser there exists a solution in the linear span of the data points. This is the core of kernel methods in machine learning as it makes the problem computationally tractable. Necessary and sufficient conditions for differentiable regularisers on Hilbert spaces to admit a representer theorem have been proved. We extend those results to nondifferentiable regularisers on uniformly convex and uniformly smooth Banach spaces. This gives a (more) complete answer to the question when there is a representer theorem. We then note that for regularised interpolation in fact the solution is determined by the function space alone and independent of the regulariser, making the extension to Banach spaces even more valuable. |
Reinforcement learning for port-Hamiltonian systems | Passivity-based control (PBC) for port-Hamiltonian systems provides an intuitive way of achieving stabilization by rendering a system passive with respect to a desired storage function. However, in most instances the control law is obtained without any performance considerations and it has to be calculated by solving a complex partial differential equation (PDE). In order to address these issues we introduce a reinforcement learning approach into the energy-balancing passivity-based control (EB-PBC) method, which is a form of PBC in which the closed-loop energy is equal to the difference between the stored and supplied energies. We propose a technique to parameterize EB-PBC that preserves the systems's PDE matching conditions, does not require the specification of a global desired Hamiltonian, includes performance criteria, and is robust to extra non-linearities such as control input saturation. The parameters of the control law are found using actor-critic reinforcement learning, enabling learning near-optimal control policies satisfying a desired closed-loop energy landscape. The advantages are that near-optimal controllers can be generated using standard energy shaping techniques and that the solutions learned can be interpreted in terms of energy shaping and damping injection, which makes it possible to numerically assess stability using passivity theory. From the reinforcement learning perspective, our proposal allows for the class of port-Hamiltonian systems to be incorporated in the actor-critic framework, speeding up the learning thanks to the resulting parameterization of the policy. The method has been successfully applied to the pendulum swing-up problem in simulations and real-life experiments. |
ROPE: Reading Order Equivariant Positional Encoding for Graph-based Document Information Extraction | Natural reading orders of words are crucial for information extraction from form-like documents. Despite recent advances in Graph Convolutional Networks (GCNs) on modeling spatial layout patterns of documents, they have limited ability to capture reading orders of given word-level node representations in a graph. We propose Reading Order Equivariant Positional Encoding (ROPE), a new positional encoding technique designed to apprehend the sequential presentation of words in documents. ROPE generates unique reading order codes for neighboring words relative to the target word given a word-level graph connectivity. We study two fundamental document entity extraction tasks including word labeling and word grouping on the public FUNSD dataset and a large-scale payment dataset. We show that ROPE consistently improves existing GCNs with a margin up to 8.4% F1-score. |
Bridge-Tower: Building Bridges Between Encoders in Vision-Language Representation Learning | Vision-Language (VL) models with the Two-Tower architecture have dominated visual-language representation learning in recent years. Current VL models either use lightweight uni-modal encoders and learn to extract, align and fuse both modalities simultaneously in a cross-modal encoder, or feed the last-layer uni-modal features directly into the top cross-modal encoder, ignoring the semantic information at the different levels in the deep uni-modal encoders. Both approaches possibly restrict vision-language representation learning and limit model performance. In this paper, we introduce multiple bridge layers that build a connection between the top layers of uni-modal encoders and each layer of the cross-modal encoder. This enables comprehensive bottom-up interactions between visual and textual representations at different semantic levels, resulting in more effective cross-modal alignment and fusion. Our proposed Bridge-Tower, pre-trained with only $4$M images, achieves state-of-the-art performance on various downstream vision-language tasks. On the VQAv2 test-std set, Bridge-Tower achieves an accuracy of $78.73\%$, outperforming the previous state-of-the-art METER model by $1.09\%$ with the same pre-training data and almost no additional parameters and computational cost. Notably, when further scaling the model, Bridge-Tower achieves an accuracy of $81.15\%$, surpassing models that are pre-trained on orders-of-magnitude larger datasets. Code is available at https://github.com/microsoft/BridgeTower. |
Neo-GNNs: Neighborhood Overlap-aware Graph Neural Networks for Link Prediction | Graph Neural Networks (GNNs) have been widely applied to various fields for learning over graph-structured data. They have shown significant improvements over traditional heuristic methods in various tasks such as node classification and graph classification. However, since GNNs heavily rely on smoothed node features rather than graph structure, they often show poor performance than simple heuristic methods in link prediction where the structural information, e.g., overlapped neighborhoods, degrees, and shortest paths, is crucial. To address this limitation, we propose Neighborhood Overlap-aware Graph Neural Networks (Neo-GNNs) that learn useful structural features from an adjacency matrix and estimate overlapped neighborhoods for link prediction. Our Neo-GNNs generalize neighborhood overlap-based heuristic methods and handle overlapped multi-hop neighborhoods. Our extensive experiments on Open Graph Benchmark datasets (OGB) demonstrate that Neo-GNNs consistently achieve state-of-the-art performance in link prediction. Our code is publicly available at https://github.com/seongjunyun/Neo_GNNs. |
Generalizing Graph Convolutional Neural Networks with Edge-Variant Recursions on Graphs | This paper reviews graph convolutional neural networks (GCNNs) through the lens of edge-variant graph filters. The edge-variant graph filter is a finite order, linear, and local recursion that allows each node, in each iteration, to weigh differently the information of its neighbors. By exploiting this recursion, we formulate a general framework for GCNNs which considers state-of-the-art solutions as particular cases. This framework results useful to i) understand the tradeoff between local detail and the number of parameters of each solution and ii) provide guidelines for developing a myriad of novel approaches that can be implemented locally in the vertex domain. One of such approaches is presented here showing superior performance w.r.t. current alternatives in graph signal classification problems. |
SpeechBrain: A General-Purpose Speech Toolkit | SpeechBrain is an open-source and all-in-one speech toolkit. It is designed to facilitate the research and development of neural speech processing technologies by being simple, flexible, user-friendly, and well-documented. This paper describes the core architecture designed to support several tasks of common interest, allowing users to naturally conceive, compare and share novel speech processing pipelines. SpeechBrain achieves competitive or state-of-the-art performance in a wide range of speech benchmarks. It also provides training recipes, pretrained models, and inference scripts for popular speech datasets, as well as tutorials which allow anyone with basic Python proficiency to familiarize themselves with speech technologies. |
Learning Mixtures of Plackett-Luce Models from Structured Partial Orders | Mixtures of ranking models have been widely used for heterogeneous preferences. However, learning a mixture model is highly nontrivial, especially when the dataset consists of partial orders. In such cases, the parameter of the model may not be even identifiable. In this paper, we focus on three popular structures of partial orders: ranked top-$l_1$, $l_2$-way, and choice data over a subset of alternatives. We prove that when the dataset consists of combinations of ranked top-$l_1$ and $l_2$-way (or choice data over up to $l_2$ alternatives), mixture of $k$ Plackett-Luce models is not identifiable when $l_1+l_2\le 2k-1$ ($l_2$ is set to $1$ when there are no $l_2$-way orders). We also prove that under some combinations, including ranked top-$3$, ranked top-$2$ plus $2$-way, and choice data over up to $4$ alternatives, mixtures of two Plackett-Luce models are identifiable. Guided by our theoretical results, we propose efficient generalized method of moments (GMM) algorithms to learn mixtures of two Plackett-Luce models, which are proven consistent. Our experiments demonstrate the efficacy of our algorithms. Moreover, we show that when full rankings are available, learning from different marginal events (partial orders) provides tradeoffs between statistical efficiency and computational efficiency. |
Adversarial Infidelity Learning for Model Interpretation | Model interpretation is essential in data mining and knowledge discovery. It can help understand the intrinsic model working mechanism and check if the model has undesired characteristics. A popular way of performing model interpretation is Instance-wise Feature Selection (IFS), which provides an importance score of each feature representing the data samples to explain how the model generates the specific output. In this paper, we propose a Model-agnostic Effective Efficient Direct (MEED) IFS framework for model interpretation, mitigating concerns about sanity, combinatorial shortcuts, model identifiability, and information transmission. Also, we focus on the following setting: using selected features to directly predict the output of the given model, which serves as a primary evaluation metric for model-interpretation methods. Apart from the features, we involve the output of the given model as an additional input to learn an explainer based on more accurate information. To learn the explainer, besides fidelity, we propose an Adversarial Infidelity Learning (AIL) mechanism to boost the explanation learning by screening relatively unimportant features. Through theoretical and experimental analysis, we show that our AIL mechanism can help learn the desired conditional distribution between selected features and targets. Moreover, we extend our framework by integrating efficient interpretation methods as proper priors to provide a warm start. Comprehensive empirical evaluation results are provided by quantitative metrics and human evaluation to demonstrate the effectiveness and superiority of our proposed method. Our code is publicly available online at https://github.com/langlrsw/MEED. |
SeqDialN: Sequential Visual Dialog Networks in Joint Visual-Linguistic Representation Space | In this work, we formulate a visual dialog as an information flow in which each piece of information is encoded with the joint visual-linguistic representation of a single dialog round. Based on this formulation, we consider the visual dialog task as a sequence problem consisting of ordered visual-linguistic vectors. For featurization, we use a Dense Symmetric Co-Attention network as a lightweight vison-language joint representation generator to fuse multimodal features (i.e., image and text), yielding better computation and data efficiencies. For inference, we propose two Sequential Dialog Networks (SeqDialN): the first uses LSTM for information propagation (IP) and the second uses a modified Transformer for multi-step reasoning (MR). Our architecture separates the complexity of multimodal feature fusion from that of inference, which allows simpler design of the inference engine. IP based SeqDialN is our baseline with a simple 2-layer LSTM design that achieves decent performance. MR based SeqDialN, on the other hand, recurrently refines the semantic question/history representations through the self-attention stack of Transformer and produces promising results on the visual dialog task. On VisDial v1.0 test-std dataset, our best single generative SeqDialN achieves 62.54% NDCG and 48.63% MRR; our ensemble generative SeqDialN achieves 63.78% NDCG and 49.98% MRR, which set a new state-of-the-art generative visual dialog model. We fine-tune discriminative SeqDialN with dense annotations and boost the performance up to 72.41% NDCG and 55.11% MRR. In this work, we discuss the extensive experiments we have conducted to demonstrate the effectiveness of our model components. We also provide visualization for the reasoning process from the relevant conversation rounds and discuss our fine-tuning methods. Our code is available at https://github.com/xiaoxiaoheimei/SeqDialN |
C-Watcher: A Framework for Early Detection of High-Risk Neighborhoods Ahead of COVID-19 Outbreak | The novel coronavirus disease (COVID-19) has crushed daily routines and is still rampaging through the world. Existing solution for nonpharmaceutical interventions usually needs to timely and precisely select a subset of residential urban areas for containment or even quarantine, where the spatial distribution of confirmed cases has been considered as a key criterion for the subset selection. While such containment measure has successfully stopped or slowed down the spread of COVID-19 in some countries, it is criticized for being inefficient or ineffective, as the statistics of confirmed cases are usually time-delayed and coarse-grained. To tackle the issues, we propose C-Watcher, a novel data-driven framework that aims at screening every neighborhood in a target city and predicting infection risks, prior to the spread of COVID-19 from epicenters to the city. In terms of design, C-Watcher collects large-scale long-term human mobility data from Baidu Maps, then characterizes every residential neighborhood in the city using a set of features based on urban mobility patterns. Furthermore, to transfer the firsthand knowledge (witted in epicenters) to the target city before local outbreaks, we adopt a novel adversarial encoder framework to learn "city-invariant" representations from the mobility-related features for precise early detection of high-risk neighborhoods, even before any confirmed cases known, in the target city. We carried out extensive experiments on C-Watcher using the real-data records in the early stage of COVID-19 outbreaks, where the results demonstrate the efficiency and effectiveness of C-Watcher for early detection of high-risk neighborhoods from a large number of cities. |
High Quality Remote Sensing Image Super-Resolution Using Deep Memory Connected Network | Single image super-resolution is an effective way to enhance the spatial resolution of remote sensing image, which is crucial for many applications such as target detection and image classification. However, existing methods based on the neural network usually have small receptive fields and ignore the image detail. We propose a novel method named deep memory connected network (DMCN) based on a convolutional neural network to reconstruct high-quality super-resolution images. We build local and global memory connections to combine image detail with environmental information. To further reduce parameters and ease time-consuming, we propose downsampling units, shrinking the spatial size of feature maps. We test DMCN on three remote sensing datasets with different spatial resolution. Experimental results indicate that our method yields promising improvements in both accuracy and visual performance over the current state-of-the-art. |
Supervised DKRC with Images for Offline System Identification | Koopman spectral theory has provided a new perspective in the field of dynamical systems in recent years. Modern dynamical systems are becoming increasingly non-linear and complex, and there is a need for a framework to model these systems in a compact and comprehensive representation for prediction and control. The central problem in applying Koopman theory to a system of interest is that the choice of finite-dimensional basis functions is typically done apriori, using expert knowledge of the systems dynamics. Our approach learns these basis functions using a supervised learning approach where a combination of autoencoders and deep neural networks learn the basis functions for any given system. We demonstrate this approach on a simple pendulum example in which we obtain a linear representation of the non-linear system and then predict the future state trajectories given some initial conditions. We also explore how changing the input representation of the dynamic systems time series data can impact the quality of learned basis functions. This alternative representation is compared to the traditional raw time series data approach to determine which method results in lower reconstruction and prediction error of the true non-linear dynamics of the system. |
Topological Deep Learning: Classification Neural Networks | Topological deep learning is a formalism that is aimed at introducing topological language to deep learning for the purpose of utilizing the minimal mathematical structures to formalize problems that arise in a generic deep learning problem. This is the first of a sequence of articles with the purpose of introducing and studying this formalism. In this article, we define and study the classification problem in machine learning in a topological setting. Using this topological framework, we show when the classification problem is possible or not possible in the context of neural networks. Finally, we demonstrate how our topological setting immediately illuminates aspects of this problem that are not as readily apparent using traditional tools. |
Moser Flow: Divergence-based Generative Modeling on Manifolds | We are interested in learning generative models for complex geometries described via manifolds, such as spheres, tori, and other implicit surfaces. Current extensions of existing (Euclidean) generative models are restricted to specific geometries and typically suffer from high computational costs. We introduce Moser Flow (MF), a new class of generative models within the family of continuous normalizing flows (CNF). MF also produces a CNF via a solution to the change-of-variable formula, however differently from other CNF methods, its model (learned) density is parameterized as the source (prior) density minus the divergence of a neural network (NN). The divergence is a local, linear differential operator, easy to approximate and calculate on manifolds. Therefore, unlike other CNFs, MF does not require invoking or backpropagating through an ODE solver during training. Furthermore, representing the model density explicitly as the divergence of a NN rather than as a solution of an ODE facilitates learning high fidelity densities. Theoretically, we prove that MF constitutes a universal density approximator under suitable assumptions. Empirically, we demonstrate for the first time the use of flow models for sampling from general curved surfaces and achieve significant improvements in density estimation, sample quality, and training complexity over existing CNFs on challenging synthetic geometries and real-world benchmarks from the earth and climate sciences. |
Shallow Bayesian Meta Learning for Real-World Few-Shot Recognition | Current state-of-the-art few-shot learners focus on developing effective training procedures for feature representations, before using simple, e.g. nearest centroid, classifiers. In this paper, we take an orthogonal approach that is agnostic to the features used and focus exclusively on meta-learning the actual classifier layer. Specifically, we introduce MetaQDA, a Bayesian meta-learning generalization of the classic quadratic discriminant analysis. This setup has several benefits of interest to practitioners: meta-learning is fast and memory-efficient, without the need to fine-tune features. It is agnostic to the off-the-shelf features chosen and thus will continue to benefit from advances in feature representations. Empirically, it leads to robust performance in cross-domain few-shot learning and, crucially for real-world applications, it leads to better uncertainty calibration in predictions. |
Boosting the Robustness Verification of DNN by Identifying the Achilles's Heel | Deep Neural Network (DNN) is a widely used deep learning technique. How to ensure the safety of DNN-based system is a critical problem for the research and application of DNN. Robustness is an important safety property of DNN. However, existing work of verifying DNN's robustness is time-consuming and hard to scale to large-scale DNNs. In this paper, we propose a boosting method for DNN robustness verification, aiming to find counter-examples earlier. Our observation is DNN's different inputs have different possibilities of existing counter-examples around them, and the input with a small difference between the largest output value and the second largest output value tends to be the achilles's heel of the DNN. We have implemented our method and applied it on Reluplex, a state-of-the-art DNN verification tool, and four DNN attacking methods. The results of the extensive experiments on two benchmarks indicate the effectiveness of our boosting method. |
ML4PG in Computer Algebra verification | ML4PG is a machine-learning extension that provides statistical proof hints during the process of Coq/SSReflect proof development. In this paper, we use ML4PG to find proof patterns in the CoqEAL library -- a library that was devised to verify the correctness of Computer Algebra algorithms. In particular, we use ML4PG to help us in the formalisation of an efficient algorithm to compute the inverse of triangular matrices. |
Learning Disentangled Representations of Video with Missing Data | Missing data poses significant challenges while learning representations of video sequences. We present Disentangled Imputed Video autoEncoder (DIVE), a deep generative model that imputes and predicts future video frames in the presence of missing data. Specifically, DIVE introduces a missingness latent variable, disentangles the hidden video representations into static and dynamic appearance, pose, and missingness factors for each object. DIVE imputes each object's trajectory where data is missing. On a moving MNIST dataset with various missing scenarios, DIVE outperforms the state of the art baselines by a substantial margin. We also present comparisons for real-world MOTSChallenge pedestrian dataset, which demonstrates the practical value of our method in a more realistic setting. Our code and data can be found at https://github.com/Rose-STL-Lab/DIVE. |
Differentially Private Coordinate Descent for Composite Empirical Risk Minimization | Machine learning models can leak information about the data used to train them. To mitigate this issue, Differentially Private (DP) variants of optimization algorithms like Stochastic Gradient Descent (DP-SGD) have been designed to trade-off utility for privacy in Empirical Risk Minimization (ERM) problems. In this paper, we propose Differentially Private proximal Coordinate Descent (DP-CD), a new method to solve composite DP-ERM problems. We derive utility guarantees through a novel theoretical analysis of inexact coordinate descent. Our results show that, thanks to larger step sizes, DP-CD can exploit imbalance in gradient coordinates to outperform DP-SGD. We also prove new lower bounds for composite DP-ERM under coordinate-wise regularity assumptions, that are nearly matched by DP-CD. For practical implementations, we propose to clip gradients using coordinate-wise thresholds that emerge from our theory, avoiding costly hyperparameter tuning. Experiments on real and synthetic data support our results, and show that DP-CD compares favorably with DP-SGD. |
Compressed Counting | Counting is among the most fundamental operations in computing. For example, counting the pth frequency moment has been a very active area of research, in theoretical computer science, databases, and data mining. When p=1, the task (i.e., counting the sum) can be accomplished using a simple counter. Compressed Counting (CC) is proposed for efficiently computing the pth frequency moment of a data stream signal A_t, where 0<p<=2. CC is applicable if the streaming data follow the Turnstile model, with the restriction that at the time t for the evaluation, A_t[i]>= 0, which includes the strict Turnstile model as a special case. For natural data streams encountered in practice, this restriction is minor. The underly technique for CC is what we call skewed stable random projections, which captures the intuition that, when p=1 a simple counter suffices, and when p = 1+/\Delta with small \Delta, the sample complexity of a counter system should be low (continuously as a function of \Delta). We show at small \Delta the sample complexity (number of projections) k = O(1/\epsilon) instead of O(1/\epsilon^2). Compressed Counting can serve a basic building block for other tasks in statistics and computing, for example, estimation entropies of data streams, parameter estimations using the method of moments and maximum likelihood. Finally, another contribution is an algorithm for approximating the logarithmic norm, \sum_{i=1}^D\log A_t[i], and logarithmic distance. The logarithmic distance is useful in machine learning practice with heavy-tailed data. |
Supervised Dictionary Learning with Auxiliary Covariates | Supervised dictionary learning (SDL) is a classical machine learning method that simultaneously seeks feature extraction and classification tasks, which are not necessarily a priori aligned objectives. The goal of SDL is to learn a class-discriminative dictionary, which is a set of latent feature vectors that can well-explain both the features as well as labels of observed data. In this paper, we provide a systematic study of SDL, including the theory, algorithm, and applications of SDL. First, we provide a novel framework that `lifts' SDL as a convex problem in a combined factor space and propose a low-rank projected gradient descent algorithm that converges exponentially to the global minimizer of the objective. We also formulate generative models of SDL and provide global estimation guarantees of the true parameters depending on the hyperparameter regime. Second, viewed as a nonconvex constrained optimization problem, we provided an efficient block coordinate descent algorithm for SDL that is guaranteed to find an $\varepsilon$-stationary point of the objective in $O(\varepsilon^{-1}(\log \varepsilon^{-1})^{2})$ iterations. For the corresponding generative model, we establish a novel non-asymptotic local consistency result for constrained and regularized maximum likelihood estimation problems, which may be of independent interest. Third, we apply SDL for imbalanced document classification by supervised topic modeling and also for pneumonia detection from chest X-ray images. We also provide simulation studies to demonstrate that SDL becomes more effective when there is a discrepancy between the best reconstructive and the best discriminative dictionaries. |
Probably certifiably correct k-means clustering | Recently, Bandeira [arXiv:1509.00824] introduced a new type of algorithm (the so-called probably certifiably correct algorithm) that combines fast solvers with the optimality certificates provided by convex relaxations. In this paper, we devise such an algorithm for the problem of k-means clustering. First, we prove that Peng and Wei's semidefinite relaxation of k-means is tight with high probability under a distribution of planted clusters called the stochastic ball model. Our proof follows from a new dual certificate for integral solutions of this semidefinite program. Next, we show how to test the optimality of a proposed k-means solution using this dual certificate in quasilinear time. Finally, we analyze a version of spectral clustering from Peng and Wei that is designed to solve k-means in the case of two clusters. In particular, we show that this quasilinear-time method typically recovers planted clusters under the stochastic ball model. |
Kalman meets Bellman: Improving Policy Evaluation through Value Tracking | Policy evaluation is a key process in Reinforcement Learning (RL). It assesses a given policy by estimating the corresponding value function. When using parameterized value functions, common approaches minimize the sum of squared Bellman temporal-difference errors and receive a point-estimate for the parameters. Kalman-based and Gaussian-processes based frameworks were suggested to evaluate the policy by treating the value as a random variable. These frameworks can learn uncertainties over the value parameters and exploit them for policy exploration. When adopting these frameworks to solve deep RL tasks, several limitations are revealed: excessive computations in each optimization step, difficulty with handling batches of samples which slows training and the effect of memory in stochastic environments which prevents off-policy learning. In this work, we discuss these limitations and propose to overcome them by an alternative general framework, based on the extended Kalman filter. We devise an optimization method, called Kalman Optimization for Value Approximation (KOVA) that can be incorporated as a policy evaluation component in policy optimization algorithms. KOVA minimizes a regularized objective function that concerns both parameter and noisy return uncertainties. We analyze the properties of KOVA and present its performance on deep RL control tasks. |
Joint Learning of Word and Label Embeddings for Sequence Labelling in Spoken Language Understanding | We propose an architecture to jointly learn word and label embeddings for slot filling in spoken language understanding. The proposed approach encodes labels using a combination of word embeddings and straightforward word-label association from the training data. Compared to the state-of-the-art methods, our approach does not require label embeddings as part of the input and therefore lends itself nicely to a wide range of model architectures. In addition, our architecture computes contextual distances between words and labels to avoid adding contextual windows, thus reducing memory footprint. We validate the approach on established spoken dialogue datasets and show that it can achieve state-of-the-art performance with much fewer trainable parameters. |
Identification of hydrodynamic instability by convolutional neural networks | The onset of hydrodynamic instabilities is of great importance in both industry and daily life, due to the dramatic mechanical and thermodynamic changes for different types of flow motions. In this paper, modern machine learning techniques, especially the convolutional neural networks (CNN), are applied to identify the transition between different flow motions raised by hydrodynamic instability, as well as critical non-dimensionalized parameters for characterizing this transit. CNN not only correctly predicts the critical transition values for both Taylor-Couette (TC) flow and Rayleigh- B\'enard (RB) convection under various setups and conditions, but also shows an outstanding performance on robustness and noise-tolerance. In addition, key spatial features used for classifying different flow patterns are revealed by the principal component analysis. |
Explaining Outcomes of Multi-Party Dialogues using Causal Learning | Multi-party dialogues are common in enterprise social media on technical as well as non-technical topics. The outcome of a conversation may be positive or negative. It is important to analyze why a dialogue ends with a particular sentiment from the point of view of conflict analysis as well as future collaboration design. We propose an explainable time series mining algorithm for such analysis. A dialogue is represented as an attributed time series of occurrences of keywords, EMPATH categories, and inferred sentiments at various points in its progress. A special decision tree, with decision metrics that take into account temporal relationships between dialogue events, is used for predicting the cause of the outcome sentiment. Interpretable rules mined from the classifier are used to explain the prediction. Experimental results are presented for the enterprise social media posts in a large company. |
On the Existence of the Adversarial Bayes Classifier (Extended Version) | Adversarial robustness is a critical property in a variety of modern machine learning applications. While it has been the subject of several recent theoretical studies, many important questions related to adversarial robustness are still open. In this work, we study a fundamental question regarding Bayes optimality for adversarial robustness. We provide general sufficient conditions under which the existence of a Bayes optimal classifier can be guaranteed for adversarial robustness. Our results can provide a useful tool for a subsequent study of surrogate losses in adversarial robustness and their consistency properties. This manuscript is the extended version of the paper "On the Existence of the Adversarial Bayes Classifier" published in NeurIPS. The results of the original paper did not apply to some non-strictly convex norms. Here we extend our results to all possible norms. Additionally, we clarify a missing step in one of our proofs. |
A Simple Convolutional Generative Network for Next Item Recommendation | Convolutional Neural Networks (CNNs) have been recently introduced in the domain of session-based next item recommendation. An ordered collection of past items the user has interacted with in a session (or sequence) are embedded into a 2-dimensional latent matrix, and treated as an image. The convolution and pooling operations are then applied to the mapped item embeddings. In this paper, we first examine the typical session-based CNN recommender and show that both the generative model and network architecture are suboptimal when modeling long-range dependencies in the item sequence. To address the issues, we introduce a simple, but very effective generative model that is capable of learning high-level representation from both short- and long-range item dependencies. The network architecture of the proposed model is formed of a stack of \emph{holed} convolutional layers, which can efficiently increase the receptive fields without relying on the pooling operation. Another contribution is the effective use of residual block structure in recommender systems, which can ease the optimization for much deeper networks. The proposed generative model attains state-of-the-art accuracy with less training time in the next item recommendation task. It accordingly can be used as a powerful recommendation baseline to beat in future, especially when there are long sequences of user feedback. |
Naive Bayes Entrapment Detection for Planetary Rovers | Entrapment detection is a prerequisite for planetary rovers to perform autonomous rescue procedure. In this study, rover entrapment and approximated entrapment criteria are formally defined. Entrapment detection using Naive Bayes classifiers is proposed and discussed along with results from experiments where the Naive Bayes entrapment detector is applied to AutoKralwer rovers. And final conclusions and further discussions are presented in the final section. |
Personalized Benchmarking with the Ludwig Benchmarking Toolkit | The rapid proliferation of machine learning models across domains and deployment settings has given rise to various communities (e.g. industry practitioners) which seek to benchmark models across tasks and objectives of personal value. Unfortunately, these users cannot use standard benchmark results to perform such value-driven comparisons as traditional benchmarks evaluate models on a single objective (e.g. average accuracy) and fail to facilitate a standardized training framework that controls for confounding variables (e.g. computational budget), making fair comparisons difficult. To address these challenges, we introduce the open-source Ludwig Benchmarking Toolkit (LBT), a personalized benchmarking toolkit for running end-to-end benchmark studies (from hyperparameter optimization to evaluation) across an easily extensible set of tasks, deep learning models, datasets and evaluation metrics. LBT provides a configurable interface for controlling training and customizing evaluation, a standardized training framework for eliminating confounding variables, and support for multi-objective evaluation. We demonstrate how LBT can be used to create personalized benchmark studies with a large-scale comparative analysis for text classification across 7 models and 9 datasets. We explore the trade-offs between inference latency and performance, relationships between dataset attributes and performance, and the effects of pretraining on convergence and robustness, showing how LBT can be used to satisfy various benchmarking objectives. |
The Advantage of Conditional Meta-Learning for Biased Regularization and Fine-Tuning | Biased regularization and fine-tuning are two recent meta-learning approaches. They have been shown to be effective to tackle distributions of tasks, in which the tasks' target vectors are all close to a common meta-parameter vector. However, these methods may perform poorly on heterogeneous environments of tasks, where the complexity of the tasks' distribution cannot be captured by a single meta-parameter vector. We address this limitation by conditional meta-learning, inferring a conditioning function mapping task's side information into a meta-parameter vector that is appropriate for that task at hand. We characterize properties of the environment under which the conditional approach brings a substantial advantage over standard meta-learning and we highlight examples of environments, such as those with multiple clusters, satisfying these properties. We then propose a convex meta-algorithm providing a comparable advantage also in practice. Numerical experiments confirm our theoretical findings. |
Deep Radar Waveform Design for Efficient Automotive Radar Sensing | In radar systems, unimodular (or constant-modulus) waveform design plays an important role in achieving better clutter/interference rejection, as well as a more accurate estimation of the target parameters. The design of such sequences has been studied widely in the last few decades, with most design algorithms requiring sophisticated a priori knowledge of environmental parameters which may be difficult to obtain in real-time scenarios. In this paper, we propose a novel hybrid model-driven and data-driven architecture that adapts to the ever changing environment and allows for adaptive unimodular waveform design. In particular, the approach lays the groundwork for developing extremely low-cost waveform design and processing frameworks for radar systems deployed in autonomous vehicles. The proposed model-based deep architecture imitates a well-known unimodular signal design algorithm in its structure, and can quickly infer statistical information from the environment using the observed data. Our numerical experiments portray the advantages of using the proposed method for efficient radar waveform design in time-varying environments. |
Deeply Supervised Active Learning for Finger Bones Segmentation | Segmentation is a prerequisite yet challenging task for medical image analysis. In this paper, we introduce a novel deeply supervised active learning approach for finger bones segmentation. The proposed architecture is fine-tuned in an iterative and incremental learning manner. In each step, the deep supervision mechanism guides the learning process of hidden layers and selects samples to be labeled. Extensive experiments demonstrated that our method achieves competitive segmentation results using less labeled samples as compared with full annotation. |
Universal Regular Conditional Distributions | We introduce a general framework for approximating regular conditional distributions (RCDs). Our approximations of these RCDs are implemented by a new class of geometric deep learning models with inputs in $\mathbb{R}^d$ and outputs in the Wasserstein-$1$ space $\mathcal{P}_1(\mathbb{R}^D)$. We find that the models built using our framework can approximate any continuous functions from $\mathbb{R}^d$ to $\mathcal{P}_1(\mathbb{R}^D)$ uniformly on compacts, and quantitative rates are obtained. We identify two methods for avoiding the "curse of dimensionality"; i.e.: the number of parameters determining the approximating neural network depends only polynomially on the involved dimension and the approximation error. The first solution describes functions in $C(\mathbb{R}^d,\mathcal{P}_1(\mathbb{R}^D))$ which can be efficiently approximated on any compact subset of $\mathbb{R}^d$. Conversely, the second approach describes sets in $\mathbb{R}^d$, on which any function in $C(\mathbb{R}^d,\mathcal{P}_1(\mathbb{R}^D))$ can be efficiently approximated. Our framework is used to obtain an affirmative answer to the open conjecture of Bishop (1994); namely: mixture density networks are universal regular conditional distributions. The predictive performance of the proposed models is evaluated against comparable learning models on various probabilistic predictions tasks in the context of ELMs, model uncertainty, and heteroscedastic regression. All the results are obtained for more general input and output spaces and thus apply to geometric deep learning contexts. |
One Objective for All Models -- Self-supervised Learning for Topic Models | Self-supervised learning has significantly improved the performance of many NLP tasks. In this paper, we highlight a key advantage of self-supervised learning -- when applied to data generated by topic models, self-supervised learning can be oblivious to the specific model, and hence is less susceptible to model misspecification. In particular, we prove that commonly used self-supervised objectives based on reconstruction or contrastive samples can both recover useful posterior information for general topic models. Empirically, we show that the same objectives can perform competitively against posterior inference using the correct model, while outperforming posterior inference using misspecified model. |
A novel auction system for selecting advertisements in Real-Time bidding | Real-Time Bidding is a new Internet advertising system that has become very popular in recent years. This system works like a global auction where advertisers bid to display their impressions in the publishers' ad slots. The most popular system to select which advertiser wins each auction is the Generalized second-price auction in which the advertiser that offers the most wins the bet and is charged with the price of the second largest bet. In this paper, we propose an alternative betting system with a new approach that not only considers the economic aspect but also other relevant factors for the functioning of the advertising system. The factors that we consider are, among others, the benefit that can be given to each advertiser, the probability of conversion from the advertisement, the probability that the visit is fraudulent, how balanced are the networks participating in RTB and if the advertisers are not paying over the market price. In addition, we propose a methodology based on genetic algorithms to optimize the selection of each advertiser. We also conducted some experiments to compare the performance of the proposed model with the famous Generalized Second-Price method. We think that this new approach, which considers more relevant aspects besides the price, offers greater benefits for RTB networks in the medium and long-term. |
Role of Data Augmentation Strategies in Knowledge Distillation for Wearable Sensor Data | Deep neural networks are parametrized by several thousands or millions of parameters, and have shown tremendous success in many classification problems. However, the large number of parameters makes it difficult to integrate these models into edge devices such as smartphones and wearable devices. To address this problem, knowledge distillation (KD) has been widely employed, that uses a pre-trained high capacity network to train a much smaller network, suitable for edge devices. In this paper, for the first time, we study the applicability and challenges of using KD for time-series data for wearable devices. Successful application of KD requires specific choices of data augmentation methods during training. However, it is not yet known if there exists a coherent strategy for choosing an augmentation approach during KD. In this paper, we report the results of a detailed study that compares and contrasts various common choices and some hybrid data augmentation strategies in KD based human activity analysis. Research in this area is often limited as there are not many comprehensive databases available in the public domain from wearable devices. Our study considers databases from small scale publicly available to one derived from a large scale interventional study into human activity and sedentary behavior. We find that the choice of data augmentation techniques during KD have a variable level of impact on end performance, and find that the optimal network choice as well as data augmentation strategies are specific to a dataset at hand. However, we also conclude with a general set of recommendations that can provide a strong baseline performance across databases. |
Filtrage vaste marge pour l'\'etiquetage s\'equentiel \`a noyaux de signaux | We address in this paper the problem of multi-channel signal sequence labeling. In particular, we consider the problem where the signals are contaminated by noise or may present some dephasing with respect to their labels. For that, we propose to jointly learn a SVM sample classifier with a temporal filtering of the channels. This will lead to a large margin filtering that is adapted to the specificity of each channel (noise and time-lag). We derive algorithms to solve the optimization problem and we discuss different filter regularizations for automated scaling or selection of channels. Our approach is tested on a non-linear toy example and on a BCI dataset. Results show that the classification performance on these problems can be improved by learning a large margin filtering. |
Algorithmic Bio-surveillance For Precise Spatio-temporal Prediction of Zoonotic Emergence | Viral zoonoses have emerged as the key drivers of recent pandemics. Human infection by zoonotic viruses are either spillover events -- isolated infections that fail to cause a widespread contagion -- or species jumps, where successful adaptation to the new host leads to a pandemic. Despite expensive bio-surveillance efforts, historically emergence response has been reactive, and post-hoc. Here we use machine inference to demonstrate a high accuracy predictive bio-surveillance capability, designed to pro-actively localize an impending species jump via automated interrogation of massive sequence databases of viral proteins. Our results suggest that a jump might not purely be the result of an isolated unfortunate cross-infection localized in space and time; there are subtle yet detectable patterns of genotypic changes accumulating in the global viral population leading up to emergence. Using tens of thousands of protein sequences simultaneously, we train models that track maximum achievable accuracy for disambiguating host tropism from the primary structure of surface proteins, and show that the inverse classification accuracy is a quantitative indicator of jump risk. We validate our claim in the context of the 2009 swine flu outbreak, and the 2004 emergence of H5N1 subspecies of Influenza A from avian reservoirs; illustrating that interrogation of the global viral population can unambiguously track a near monotonic risk elevation over several preceding years leading to eventual emergence. |
Confident Kernel Sparse Coding and Dictionary Learning | In recent years, kernel-based sparse coding (K-SRC) has received particular attention due to its efficient representation of nonlinear data structures in the feature space. Nevertheless, the existing K-SRC methods suffer from the lack of consistency between their training and test optimization frameworks. In this work, we propose a novel confident K-SRC and dictionary learning algorithm (CKSC) which focuses on the discriminative reconstruction of the data based on its representation in the kernel space. CKSC focuses on reconstructing each data sample via weighted contributions which are confident in its corresponding class of data. We employ novel discriminative terms to apply this scheme to both training and test frameworks in our algorithm. This specific design increases the consistency of these optimization frameworks and improves the discriminative performance in the recall phase. In addition, CKSC directly employs the supervised information in its dictionary learning framework to enhance the discriminative structure of the dictionary. For empirical evaluations, we implement our CKSC algorithm on multivariate time-series benchmarks such as DynTex++ and UTKinect. Our claims regarding the superior performance of the proposed algorithm are justified throughout comparing its classification results to the state-of-the-art K-SRC algorithms. |
FedSiam: Towards Adaptive Federated Semi-Supervised Learning | Federated learning (FL) has emerged as an effective technique to co-training machine learning models without actually sharing data and leaking privacy. However, most existing FL methods focus on the supervised setting and ignore the utilization of unlabeled data. Although there are a few existing studies trying to incorporate unlabeled data into FL, they all fail to maintain performance guarantees or generalization ability in various real-world settings. In this paper, we focus on designing a general framework FedSiam to tackle different scenarios of federated semi-supervised learning, including four settings in the labels-at-client scenario and two setting in the labels-at-server scenario. FedSiam is built upon a siamese network into FL with a momentum update to handle the non-IID challenges introduced by unlabeled data. We further propose a new metric to measure the divergence of local model layers within the siamese network. Based on the divergence, FedSiam can automatically select layer-level parameters to be uploaded to the server in an adaptive manner. Experimental results on three datasets under two scenarios with different data distribution settings demonstrate that the proposed FedSiam framework outperforms state-of-the-art baselines. |
On the Intrinsic Robustness of NVM Crossbars Against Adversarial Attacks | The increasing computational demand of Deep Learning has propelled research in special-purpose inference accelerators based on emerging non-volatile memory (NVM) technologies. Such NVM crossbars promise fast and energy-efficient in-situ Matrix Vector Multiplication (MVM) thus alleviating the long-standing von Neuman bottleneck in today's digital hardware. However, the analog nature of computing in these crossbars is inherently approximate and results in deviations from ideal output values, which reduces the overall performance of Deep Neural Networks (DNNs) under normal circumstances. In this paper, we study the impact of these non-idealities under adversarial circumstances. We show that the non-ideal behavior of analog computing lowers the effectiveness of adversarial attacks, in both Black-Box and White-Box attack scenarios. In a non-adaptive attack, where the attacker is unaware of the analog hardware, we observe that analog computing offers a varying degree of intrinsic robustness, with a peak adversarial accuracy improvement of 35.34%, 22.69%, and 9.90% for white box PGD (epsilon=1/255, iter=30) for CIFAR-10, CIFAR-100, and ImageNet respectively. We also demonstrate "Hardware-in-Loop" adaptive attacks that circumvent this robustness by utilizing the knowledge of the NVM model. |
Facts as Experts: Adaptable and Interpretable Neural Memory over Symbolic Knowledge | Massive language models are the core of modern NLP modeling and have been shown to encode impressive amounts of commonsense and factual information. However, that knowledge exists only within the latent parameters of the model, inaccessible to inspection and interpretation, and even worse, factual information memorized from the training corpora is likely to become stale as the world changes. Knowledge stored as parameters will also inevitably exhibit all of the biases inherent in the source materials. To address these problems, we develop a neural language model that includes an explicit interface between symbolically interpretable factual information and subsymbolic neural knowledge. We show that this model dramatically improves performance on two knowledge-intensive question-answering tasks. More interestingly, the model can be updated without re-training by manipulating its symbolic representations. In particular this model allows us to add new facts and overwrite existing ones in ways that are not possible for earlier models. |
Reconstructing Speech Stimuli From Human Auditory Cortex Activity Using a WaveNet Approach | The superior temporal gyrus (STG) region of cortex critically contributes to speech recognition. In this work, we show that a proposed WaveNet, with limited available data, is able to reconstruct speech stimuli from STG intracranial recordings. We further investigate the impulse response of the fitted model for each recording electrode and observe phoneme level temporospectral tuning properties for the recorded area of cortex. This discovery is consistent with previous studies implicating the posterior STG (pSTG) in a phonetic representation of speech and provides detailed acoustic features that certain electrode sites possibly extract during speech recognition. |
A Kernel Loss for Solving the Bellman Equation | Value function learning plays a central role in many state-of-the-art reinforcement-learning algorithms. Many popular algorithms like Q-learning do not optimize any objective function, but are fixed-point iterations of some variant of Bellman operator that is not necessarily a contraction. As a result, they may easily lose convergence guarantees, as can be observed in practice. In this paper, we propose a novel loss function, which can be optimized using standard gradient-based methods without risking divergence. The key advantage is that its gradient can be easily approximated using sampled transitions, avoiding the need for double samples required by prior algorithms like residual gradient. Our approach may be combined with general function classes such as neural networks, on either on- or off-policy data, and is shown to work reliably and effectively in several benchmarks. |
Adversarial Skill Networks: Unsupervised Robot Skill Learning from Video | Key challenges for the deployment of reinforcement learning (RL) agents in the real world are the discovery, representation and reuse of skills in the absence of a reward function. To this end, we propose a novel approach to learn a task-agnostic skill embedding space from unlabeled multi-view videos. Our method learns a general skill embedding independently from the task context by using an adversarial loss. We combine a metric learning loss, which utilizes temporal video coherence to learn a state representation, with an entropy regularized adversarial skill-transfer loss. The metric learning loss learns a disentangled representation by attracting simultaneous viewpoints of the same observations and repelling visually similar frames from temporal neighbors. The adversarial skill-transfer loss enhances re-usability of learned skill embeddings over multiple task domains. We show that the learned embedding enables training of continuous control policies to solve novel tasks that require the interpolation of previously seen skills. Our extensive evaluation with both simulation and real world data demonstrates the effectiveness of our method in learning transferable skills from unlabeled interaction videos and composing them for new tasks. Code, pretrained models and dataset are available at http://robotskills.cs.uni-freiburg.de |
Incremental Transductive Learning Approaches to Schistosomiasis Vector Classification | The key issues pertaining to collection of epidemic disease data for our analysis purposes are that it is a labour intensive, time consuming and expensive process resulting in availability of sparse sample data which we use to develop prediction models. To address this sparse data issue, we present novel Incremental Transductive methods to circumvent the data collection process by applying previously acquired data to provide consistent, confidence-based labelling alternatives to field survey research. We investigated various reasoning approaches for semisupervised machine learning including Bayesian models for labelling data. The results show that using the proposed methods, we can label instances of data with a class of vector density at a high level of confidence. By applying the Liberal and Strict Training Approaches, we provide a labelling and classification alternative to standalone algorithms. The methods in this paper are components in the process of reducing the proliferation of the Schistosomiasis disease and its effects. |
A Generalized Weisfeiler-Lehman Graph Kernel | The Weisfeiler-Lehman graph kernels are among the most prevalent graph kernels due to their remarkable time complexity and predictive performance. Their key concept is based on an implicit comparison of neighborhood representing trees with respect to equality (i.e., isomorphism). This binary valued comparison is, however, arguably too rigid for defining suitable similarity measures over graphs. To overcome this limitation, we propose a generalization of Weisfeiler-Lehman graph kernels which takes into account the similarity between trees rather than equality. We achieve this using a specifically fitted variation of the well-known tree edit distance which can efficiently be calculated. We empirically show that our approach significantly outperforms state-of-the-art methods in terms of predictive performance on datasets containing structurally more complex graphs beyond the typically considered molecular graphs. |
SemiRetro: Semi-template framework boosts deep retrosynthesis prediction | Recently, template-based (TB) and template-free (TF) molecule graph learning methods have shown promising results to retrosynthesis. TB methods are more accurate using pre-encoded reaction templates, and TF methods are more scalable by decomposing retrosynthesis into subproblems, i.e., center identification and synthon completion. To combine both advantages of TB and TF, we suggest breaking a full-template into several semi-templates and embedding them into the two-step TF framework. Since many semi-templates are reduplicative, the template redundancy can be reduced while the essential chemical knowledge is still preserved to facilitate synthon completion. We call our method SemiRetro, introduce a new GNN layer (DRGAT) to enhance center identification, and propose a novel self-correcting module to improve semi-template classification. Experimental results show that SemiRetro significantly outperforms both existing TB and TF methods. In scalability, SemiRetro covers 98.9\% data using 150 semi-templates, while previous template-based GLN requires 11,647 templates to cover 93.3\% data. In top-1 accuracy, SemiRetro exceeds template-free G2G 4.8\% (class known) and 6.0\% (class unknown). Besides, SemiRetro has better training efficiency than existing methods. |
Embracing Imperfect Datasets: A Review of Deep Learning Solutions for Medical Image Segmentation | The medical imaging literature has witnessed remarkable progress in high-performing segmentation models based on convolutional neural networks. Despite the new performance highs, the recent advanced segmentation models still require large, representative, and high quality annotated datasets. However, rarely do we have a perfect training dataset, particularly in the field of medical imaging, where data and annotations are both expensive to acquire. Recently, a large body of research has studied the problem of medical image segmentation with imperfect datasets, tackling two major dataset limitations: scarce annotations where only limited annotated data is available for training, and weak annotations where the training data has only sparse annotations, noisy annotations, or image-level annotations. In this article, we provide a detailed review of the solutions above, summarizing both the technical novelties and empirical results. We further compare the benefits and requirements of the surveyed methodologies and provide our recommended solutions. We hope this survey article increases the community awareness of the techniques that are available to handle imperfect medical image segmentation datasets. |
Self-Supervised Graph Learning with Proximity-based Views and Channel Contrast | We consider graph representation learning in a self-supervised manner. Graph neural networks (GNNs) use neighborhood aggregation as a core component that results in feature smoothing among nodes in proximity. While successful in various prediction tasks, such a paradigm falls short of capturing nodes' similarities over a long distance, which proves to be important for high-quality learning. To tackle this problem, we strengthen the graph with two additional graph views, in which nodes are directly linked to those with the most similar features or local structures. Not restricted by connectivity in the original graph, the generated views allow the model to enhance its expressive power with new and complementary perspectives from which to look at the relationship between nodes. Following a contrastive learning approach, we propose a method that aims to maximize the agreement between representations across generated views and the original graph. We also propose a channel-level contrast approach that greatly reduces computation cost, compared to the commonly used node level contrast, which requires computation cost quadratic in the number of nodes. Extensive experiments on seven assortative graphs and four disassortative graphs demonstrate the effectiveness of our approach. |
Deep Reinforcement Learning for General Video Game AI | The General Video Game AI (GVGAI) competition and its associated software framework provides a way of benchmarking AI algorithms on a large number of games written in a domain-specific description language. While the competition has seen plenty of interest, it has so far focused on online planning, providing a forward model that allows the use of algorithms such as Monte Carlo Tree Search. In this paper, we describe how we interface GVGAI to the OpenAI Gym environment, a widely used way of connecting agents to reinforcement learning problems. Using this interface, we characterize how widely used implementations of several deep reinforcement learning algorithms fare on a number of GVGAI games. We further analyze the results to provide a first indication of the relative difficulty of these games relative to each other, and relative to those in the Arcade Learning Environment under similar conditions. |
Sequential Recommendation with Relation-Aware Kernelized Self-Attention | Recent studies identified that sequential Recommendation is improved by the attention mechanism. By following this development, we propose Relation-Aware Kernelized Self-Attention (RKSA) adopting a self-attention mechanism of the Transformer with augmentation of a probabilistic model. The original self-attention of Transformer is a deterministic measure without relation-awareness. Therefore, we introduce a latent space to the self-attention, and the latent space models the recommendation context from relation as a multivariate skew-normal distribution with a kernelized covariance matrix from co-occurrences, item characteristics, and user information. This work merges the self-attention of the Transformer and the sequential recommendation by adding a probabilistic model of the recommendation task specifics. We experimented RKSA over the benchmark datasets, and RKSA shows significant improvements compared to the recent baseline models. Also, RKSA were able to produce a latent space model that answers the reasons for recommendation. |
Markov Blanket Discovery using Minimum Message Length | Causal discovery automates the learning of causal Bayesian networks from data and has been of active interest from their beginning. With the sourcing of large data sets off the internet, interest in scaling up to very large data sets has grown. One approach to this is to parallelize search using Markov Blanket (MB) discovery as a first step, followed by a process of combining MBs in a global causal model. We develop and explore three new methods of MB discovery using Minimum Message Length (MML) and compare them empirically to the best existing methods, whether developed specifically as MB discovery or as feature selection. Our best MML method is consistently competitive and has some advantageous features. |
New Era of Deeplearning-Based Malware Intrusion Detection: The Malware Detection and Prediction Based On Deep Learning | With the development of artificial intelligence algorithms like deep learning models and the successful applications in many different fields, further similar trails of deep learning technology have been made in cyber security area. It shows the preferable performance not only in academic security research but also in industry practices when dealing with part of cyber security issues by deep learning methods compared to those conventional rules. Especially for the malware detection and classification tasks, it saves generous time cost and promotes the accuracy for a total pipeline of malware detection system. In this paper, we construct special deep neural network, ie, MalDeepNet (TB-Malnet and IB-Malnet) for malware dynamic behavior classification tasks. Then we build the family clustering algorithm based on deep learning and fulfil related testing. Except that, we also design a novel malware prediction model which could detect the malware coming in future through the Mal Generative Adversarial Network (Mal-GAN) implementation. All those algorithms present fairly considerable value in related datasets afterwards. |
Automatic Large-Scale Data Acquisition via Crowdsourcing for Crosswalk Classification: A Deep Learning Approach | Correctly identifying crosswalks is an essential task for the driving activity and mobility autonomy. Many crosswalk classification, detection and localization systems have been proposed in the literature over the years. These systems use different perspectives to tackle the crosswalk classification problem: satellite imagery, cockpit view (from the top of a car or behind the windshield), and pedestrian perspective. Most of the works in the literature are designed and evaluated using small and local datasets, i.e. datasets that present low diversity. Scaling to large datasets imposes a challenge for the annotation procedure. Moreover, there is still need for cross-database experiments in the literature because it is usually hard to collect the data in the same place and conditions of the final application. In this paper, we present a crosswalk classification system based on deep learning. For that, crowdsourcing platforms, such as OpenStreetMap and Google Street View, are exploited to enable automatic training via automatic acquisition and annotation of a large-scale database. Additionally, this work proposes a comparison study of models trained using fully-automatic data acquisition and annotation against models that were partially annotated. Cross-database experiments were also included in the experimentation to show that the proposed methods enable use with real world applications. Our results show that the model trained on the fully-automatic database achieved high overall accuracy (94.12%), and that a statistically significant improvement (to 96.30%) can be achieved by manually annotating a specific part of the database. Finally, the results of the cross-database experiments show that both models are robust to the many variations of image and scenarios, presenting a consistent behavior. |
Zero-Shot Anticipation for Instructional Activities | How can we teach a robot to predict what will happen next for an activity it has never seen before? We address this problem of zero-shot anticipation by presenting a hierarchical model that generalizes instructional knowledge from large-scale text-corpora and transfers the knowledge to the visual domain. Given a portion of an instructional video, our model predicts coherent and plausible actions multiple steps into the future, all in rich natural language. To demonstrate the anticipation capabilities of our model, we introduce the Tasty Videos dataset, a collection of 2511 recipes for zero-shot learning, recognition and anticipation. |
Machine Learning for recognition of minerals from multispectral data | Machine Learning (ML) has found several applications in spectroscopy, including being used to recognise minerals and estimate elemental composition. In this work, we present novel methods for automatic mineral identification based on combining data from different spectroscopic methods. We evaluate combining data from three spectroscopic methods: vibrational Raman scattering, reflective Visible-Near Infrared (VNIR), and Laser-Induced Breakdown Spectroscopy (LIBS). These methods were paired into Raman + VNIR, Raman + LIBS and VNIR + LIBS, and different methods of data fusion applied to each pair to classify minerals. The methods presented here are shown to outperform the use of a single data source by a significant margin. Additionally, we present a Deep Learning algorithm for mineral classification from Raman spectra that outperforms previous state-of-the-art methods. Our approach was tested on various open access experimental Raman (RRUFF) and VNIR (USGS, Relab, ECOSTRESS), as well as synthetic LIBS NIST spectral libraries. Our cross-validation tests show that multi-method spectroscopy paired with ML paves the way towards rapid and accurate characterization of rocks and minerals. |
Fairness and Accountability Design Needs for Algorithmic Support in High-Stakes Public Sector Decision-Making | Calls for heightened consideration of fairness and accountability in algorithmically-informed public decisions---like taxation, justice, and child protection---are now commonplace. How might designers support such human values? We interviewed 27 public sector machine learning practitioners across 5 OECD countries regarding challenges understanding and imbuing public values into their work. The results suggest a disconnect between organisational and institutional realities, constraints and needs, and those addressed by current research into usable, transparent and 'discrimination-aware' machine learning---absences likely to undermine practical initiatives unless addressed. We see design opportunities in this disconnect, such as in supporting the tracking of concept drift in secondary data sources, and in building usable transparency tools to identify risks and incorporate domain knowledge, aimed both at managers and at the 'street-level bureaucrats' on the frontlines of public service. We conclude by outlining ethical challenges and future directions for collaboration in these high-stakes applications. |