Skip to content

Repository files navigation

Machine Learning Algorithms From Scratch

A minimal and educational repository implementing core Machine Learning algorithms from scratch — no built-in ML libraries used.

This project is designed for learning and understanding the underlying logic, intuition, and mathematics behind machine learning models. It avoids the use of libraries like scikit-learn, xgboost, or lightgbm for model building, and focuses on building everything from the ground up using pure Python and NumPy.


Features

  • Implementations of core ML algorithms from scratch
  • Focus on interpretability and step-by-step logic
  • Uses only Python and NumPy
  • Great for beginners who want to understand how ML works under the hood

Implemented Algorithms

Regression

  • Linear Regression
  • Ridge Regression
  • K-Nearest Neighbors

Classification

  • Logistic Regression
  • Gaussian Determinant Analysis (GDA)
  • Support Vector Machine (SVM)
  • K-Nearest Neighbors

Installation & Usage

  1. Clone the repository:

    git clone https://github.com/anshul-dying/ML-Algos-From-Scratch.git
    cd ML-Algos-From-Scratch
  2. Install Dependencies:

     pip install pandas, numpy, matplotlib, scikit-learn, seaborn

Goals

  • Learn core ML principles by implementing models manually
  • Build mathematical and algorithmic intuition
  • Avoid abstraction-heavy libraries until the fundamentals are solid

Contributing

Contributions are welcome! If you want to:

  • Add a new algorithm
  • Improve code readability or efficiency
  • Fix bugs or incorrect math Please open a pull request or create an issue.

Author

Made with curiosity by Anshul

About

Machine Learning from Scratch A minimal repo implementing core ML algorithms without using any built-in ML libraries. Made purely for understanding the underlying logic and math.

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages