ML from Scratch (C)

This repository contains a small machine learning library written in plain C. The goal is to understand machine learning fundamentals by implementing them from scratch, starting with linear algebra and basic models.

Goals

• Learn ML by implementing core concepts manually
• Build a small, reusable linear algebra library
• Keep the code simple, readable, and correct

Non-goals

• Production performance
• GPU acceleration
• Competing with BLAS / existing ML frameworks

Structure

• linalg/ — basic linear algebra (matrices, operations)
• ml/ — machine learning models built on top of linalg
• examples/ — small programs using the libraries
• datasets/ — small datasets (e.g. Iris)

Run synthetic linear regression example

Build and run the example:

gcc -I./linalg/include -I./ml/include \
    linalg/src/matrix.c \
    linalg/src/linalg.c \
    ml/src/linear_regression.c \
    examples/synthetic_linear_regression.c \
    -o bin/synthetic_linear_regression

./bin/synthetic_linear_regression

Status

Early stage. APIs may change as concepts are added.

License

MIT