// the find
Avik-Jain/100-Days-Of-ML-Code
100 Days of ML Coding
A learning journal from 2018 where someone documented 100 days of studying ML concepts, mostly by watching YouTube videos and working through scikit-learn tutorials. It's not a library or tool — it's a study log with some hand-drawn infographics. For developers who want reference implementations of classical ML algorithms, the code coverage is thin; for beginners looking for a structured syllabus, the video links are mostly to channels that still exist.
The infographics for algorithms like SVM, decision trees, and random forests are genuinely well-drawn and useful as quick visual references. The scikit-learn code snippets are concise and runnable without much setup. It covers a reasonable breadth of classical ML in roughly the right pedagogical order — preprocessing, linear models, then tree-based methods, then neural network intuition. The 3Blue1Brown playlist recommendations (linear algebra, calculus, neural networks) are still among the best introductory material available.
The project stopped at day 54 out of 100, and the last commit was in 2023 with no substantive new content — it's been effectively abandoned for years. About half the 'days' are just diary entries saying 'I watched a video' with no code or notes, which adds star count but not value. The code that does exist is scattered across markdown files with no unified notebook or runnable project structure — you can't clone and run anything end-to-end. Nothing here touches modern ML: no PyTorch, no transformers, no data pipelines beyond toy CSVs.