Logistic Regression Classifier Visualization App

Demo

Here’s an example of the application interface:

This project is a Streamlit-based web application that demonstrates the working of a Logistic Regression Classifier on synthetic datasets. The app allows users to configure various hyperparameters for logistic regression, visualize the dataset, and observe the decision boundaries dynamically.

Features

• Dataset Selection: Choose between Binary or Multiclass datasets generated using make_blobs.
• Hyperparameter Tuning: Modify Logistic Regression parameters, including:
  • Regularization (penalty)
  • Regularization strength (C)
  • Solver type (solver)
  • Maximum iterations (max_iter)
  • Multi-class strategy (multi_class)
  • ElasticNet ratio (l1_ratio)
  • Tolerance for convergence (tol)
  • Class weight (class_weight)
  • Fit intercept, dual optimization, warm start, and more.
• Interactive Visualization:
  • Plot the dataset points with decision boundaries.
  • Observe how changes in hyperparameters affect classification results.
• Accuracy Calculation: View the classifier's performance with the current settings.

Requirements

To run this project, you'll need the following libraries:

• streamlit
• numpy
• matplotlib
• scikit-learn

You can install them using:

pip install streamlit numpy matplotlib scikit-learn

How to Run

1. Clone the repository:

   git clone https://github.com/yourusername/logistic-regression-visualization.git
   cd logistic-regression-visualization

2. Run the Streamlit app:

   streamlit run app.py

3. Open the URL displayed in your terminal to view the app in your browser.

File Structure

• app.py: Main Streamlit application file.
• README.md: Project documentation (this file).

How It Works

1. Dataset Creation: Synthetic datasets are generated using the make_blobs function from scikit-learn.
2. User Interaction: The sidebar allows users to modify hyperparameters of the Logistic Regression model.
3. Visualization: The decision boundary and dataset points are plotted using matplotlib.
4. Classifier Training: Logistic Regression is trained on the dataset, and predictions are made.
5. Results: The accuracy of the classifier and an updated decision boundary plot are displayed.