I focus on understanding each concept from first principles, then explaining it in my own words. I do not rely on LLMs to write code. LLMs are used only for research and conceptual clarity.
The goal is long term mastery, not shortcuts.
Foundations Core ideas of machine learning. Problem types. Model workflow. Bias variance intuition.
Math Fundamentals Mathematics required for machine learning. Linear algebra probability and statistics. Implemented from scratch using NumPy to build intuition.
Supervised Learning Exploratory Data Analysis. Regression and classification problems. End to end pipelines. Data preprocessing feature scaling model training evaluation. Each algorithm is explained conceptually before implementation.
Unsupervised Learning Clustering and dimensionality reduction. Concepts and intuition first. Implementations in progress.
Why this repo exists.
To build deep understanding. To create a personal reference. To avoid black box learning. To document learning honestly and clearly.
Status.
Actively maintained. Unsupervised section in progress.