metadata

license: apache-2.0
tags:
  - keras
  - time-series
  - lstm
  - regression
datasets:
  - output8.csv
metrics:
  - mean_squared_error
model_name: my_model

bhaskar1

The above model is a simple neural network built using TensorFlow/Keras. It is designed to perform a regression task, which means it predicts continuous numeric values based on input features.

Keras Model for Time Series Prediction

This repository contains a Keras model trained for time series prediction. The model has been deployed to Hugging Face Hub and can be used to predict numeric values based on timestamped data.

Model Overview

The model is a simple neural network built using Keras. It is designed to perform regression tasks, predicting a numeric value from input features.

Architecture

Input Layer: 10 neurons, ReLU activation
Hidden Layer: 10 neurons, ReLU activation
Output Layer: 1 neuron (for regression)

How to Use the Model

To use this model, follow the steps below:

1. Install Required Libraries

Make sure you have the necessary libraries installed:

pip install tensorflow huggingface-hub pandas matplotlib

#load the model from the Hugging Face Hub:
import tensorflow as tf
from huggingface_hub import from_pretrained_keras
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

# Load the model from Hugging Face Hub
model = from_pretrained_keras("iiitbh18/bhaskar1")

# Prepare Dummy Data

# Create a DataFrame with dummy data
dummy_data = {
    'timestamp': [
        '2024-08-07 02:43:28.788',
        '2024-08-07 02:43:28.788',
        '2024-08-07 02:43:43.788',
        '2024-08-07 02:43:43.788',
        '2024-08-07 02:43:58.788'
    ],
    'value': [
        99.00000005960464,
        98.90000000596046,
        98.70000004768372,
        99.00000005960464,
        98.89999993145466
    ]
}

df = pd.DataFrame(dummy_data)

#Prepare Data for Prediction

X_dummy = df['value'].values.reshape(-1, 1)  # Reshape to match model's input shape








# Make Predictions

predictions = model.predict(X_dummy)
print("Predictions:", predictions)

#Visualize the Results

# Plot actual vs. predicted values
plt.figure(figsize=(10, 6))
plt.scatter(df['timestamp'], df['value'], color='blue', label='Actual Values')
plt.scatter(df['timestamp'], predictions, color='red', label='Predicted Values')
plt.xlabel('Timestamp')
plt.ylabel('Value')
plt.title('Actual vs Predicted Values')
plt.legend()
plt.xticks(rotation=45)
plt.show()