ML from scratch

This repo is my way of really understanding machine learning—by building everything from scratch. No scikit-learn, no TensorFlow, just NumPy and raw math. Writing these algorithms line by line forces me to truly get how they work, rather than just using them as black boxes.

So far, I’ve implemented models like Perceptron, Softmax Regression, Gaussian Discriminant Analysis, and Stochastic Gradient Descent (SGD). Each one is built with clarity in mind, making it easier to see what’s happening under the hood.

This is an ongoing project, and I keep adding new implementations as I explore more ML concepts. If you’re into machine learning and love breaking things down to their fundamentals, you might find this repo interesting. Check it out on GitHub!