Introduction to Machine Learning 2024 from Çağan Kiper

This repository contains the Jupyter Notebooks used in the "Introduction to Machine Learning" course provided by ITU AI CLUB. These notebooks cover foundational topics and practical implementations to help beginners get started with machine learning concepts.

Table of Contents

1. Overview
2. Setup Instructions
3. Notebooks
   • Week 1: Introduction to Linear Regression
   • Week 2: Loss Functions and Optimization
   • Week 3: Regularization and Model Selection
   • Week 4: Support Vector Machines
4. Contributions
5. License

Overview

These notebooks are designed to provide students with:

• A conceptual understanding of machine learning techniques.
• Hands-on experience in implementing algorithms using Python.
• A structured progression from basic concepts to advanced methods.

Setup Instructions

To use these notebooks, follow these steps:

1. Clone the Repository:

   git clone https://github.com/yourusername/intro-to-ml-course.git
   cd intro-to-ml-course

2. Install Dependencies: Create a virtual environment and install required libraries:

   python -m venv venv
   source venv/bin/activate  # Use `venv\Scripts\activate` on Windows
   pip install -r requirements.txt

3. Start Jupyter Notebook:

   jupyter notebook

4. Open the desired notebook and start exploring!

Notebooks

Week 1: Introduction to Linear Regression

• Basics of regression analysis.
• Understanding the loss function (MSE).
• Optimization methods including the Normal Equation and Gradient Descent.

Week 2: Loss Functions and Optimization

• Detailed discussion on Mean Squared Error and its variations.
• Introduction to optimization techniques.
• Variants of Gradient Descent (Batch, Stochastic, and Mini-batch).

Week 3: Regularization and Model Selection

• Overfitting and the need for regularization.
• L1 (Lasso), L2 (Ridge), and Elastic Net techniques.
• Cross-validation and Grid Search for hyperparameter tuning.

Week 4: Support Vector Machines (SVM)

• Concepts of SVM, including hyperplane and margin.
• Linear SVM optimization.
• Non-linear SVMs and the Kernel Trick.
• Practical implementation with Soft Margin SVM.

Contributions

We welcome contributions to improve this repository. Please submit pull requests for any enhancements or fixes.

License

This project is licensed under the GNU GPL-3.0 license.

Although everything is free to use, modify and distribute, credit is always appreciated.