gradient-descent-and-logistic-regression

Overview

This project implements gradient descent optimization and logistic regression with regularization. It includes:

• Gradient descent with customizable learning rate strategies.
• Regularized logistic regression supporting L1 and L2 penalties.
• Visualizations of optimization behavior and model performance.

Files

1. modules.py: Core modules for L1/L2 regularization and logistic regression objectives.
2. gradient_descent.py: Main gradient descent implementation with fixed/exponential learning rates.
3. learning_rate.py: Learning rate strategies, including FixedLR and ExponentialLR classes.
4. logistic_regression.py: Logistic regression classifier with L1/L2 regularization and cross-validation support.
5. gradient_descent_investigation.py: Analysis tools for visualizing gradient descent behavior and comparing learning rates.