https://github.com/asut00/machine-learning-piscine_42ai

Comprehensive Machine Learning Bootcamp by 42AI: hands-on modules on regression, gradient descent, and real-world ML applications.
https://github.com/asut00/machine-learning-piscine_42ai

linear-regression machine-learning matplotlib numpy pandas python scikit-learn

Last synced: 27 days ago
JSON representation

Comprehensive Machine Learning Bootcamp by 42AI: hands-on modules on regression, gradient descent, and real-world ML applications.

• Host: GitHub
• URL: https://github.com/asut00/machine-learning-piscine_42ai
• Owner: asut00
• Created: 2024-12-17T21:32:26.000Z (29 days ago)
• Default Branch: main
• Last Pushed: 2024-12-17T22:38:20.000Z (29 days ago)
• Last Synced: 2024-12-17T23:26:32.018Z (29 days ago)
• Topics: linear-regression, machine-learning, matplotlib, numpy, pandas, python, scikit-learn
• Language: Python
• Homepage:
• Size: 3.76 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        

# Machine Learning Bootcamp - 42AI 🦾

## Overview

This repository contains all the materials, exercises, and solutions from the week-long Machine Learning Bootcamp organized by 42AI, the Artificial Intelligence association of École 42. The bootcamp was designed to provide a comprehensive introduction to Machine Learning (ML), including foundational mathematics, implementation of ML algorithms, and practical applications. Participants worked through hands-on projects to build a strong understanding of machine learning principles.

## Contents

The bootcamp is divided into several modules, each focusing on key Machine Learning concepts and techniques. Below is an outline of the modules:

### Day 1: Stepping into Machine Learning

- **Topics Covered:**

  - Introduction to Machine Learning

  - Supervised vs. Unsupervised Learning

  - Linear Algebra Review

    - Matrix operations

    - Transpose, addition, multiplication

  - Building a Regression Model

    - Hypothesis, cost function, and evaluation metrics

- **Key Exercises:**

  - Implementing a `Matrix` and `Vector` class with basic operations

  - Creating a basic statistical class, `TinyStatistician`, for computing mean, median, variance, and standard deviation

### Day 2: Univariate Linear Regression

- **Topics Covered:**

  - Gradient Descent for Linear Regression

  - Cost function minimization

  - Learning rate tuning

  - Normalization techniques

- **Key Exercises:**

  - Implementing gradient descent iteratively and using vectorization

  - Training a regression model to predict outcomes from one feature

### Day 3: Multivariate Linear Regression

- **Topics Covered:**

  - Extending regression models to handle multiple features

  - Polynomial regression for non-linear relationships

  - Feature scaling and engineering

  - Understanding overfitting and underfitting

- **Key Exercises:**

  - Implementing multivariate linear regression using vectorized gradient descent

  - Evaluating model performance with loss functions

  - Implementing polynomial feature expansion

### Day 4: Logistic Regression  

**Topics Covered:**  

- Introduction to Logistic Regression  

  - Logistic hypothesis  

  - Decision boundary and cost function  

  - Gradient descent for logistic regression  

- Classification Metrics  

  - Accuracy, precision, recall, F1-score, and confusion matrix  

**Key Exercises:**  

- Implementation of the sigmoid function and logistic hypothesis.  

- Building a logistic regression model using gradient descent.  

- Evaluating performance through classification metrics (confusion matrix, F1-score).  

### Day 5: Regularization  

**Topics Covered:**  

- Understanding Overfitting  

  - Causes and how to detect overfitting  

  - Concepts of regularization (L2 regularization)  

- Regularized Regression  

  - Regularized cost functions for linear and logistic regression  

  - Regularized gradient descent  

**Key Exercises:**  

- Implementing L2 regularization and applying it to regression models.  

- Extending linear and logistic regression to include regularization for better generalization.  

- Comparing regularized and non-regularized models to understand their impact on performance.  

## Acknowledgments

This bootcamp was organized by 42AI, with materials inspired by the Machine Learning MOOC by Andrew Ng. Special thanks to the mentors and peers who contributed to the collaborative learning environment.

## Resources

- [Andrew Ng’s Machine Learning Course on Coursera](https://www.coursera.org/learn/machine-learning)

- [42AI GitHub Repository](https://github.com/42-AI)