This is a lightly "edited" version of the free PDF of the great book by @abhishekkrthakur:
Approaching (Almost) Any Machine Learning Problem
- Added a two-level heading structure, covering the entirety of the 300 pages, easier navigation.
- Added notes for typos I've found.
- Added the original cover page, better visibility with thumbnails.
Example:
Note1: There are some personal notes for some paragraphs and highlighted sentences, but
you can remove everything easily by clicking on it and "trash bin" icon, if you just want the headers.
Note2: The pdf doesn't have any subheaders in the text, ie a) LSTM, b) Transformers etc... so my headers match
the beginning of paragraphs where the author starts addressing these topics.
Nothing in the book has been modified.
AAAMLP is a great book on fundamental applied ML practices and classic Kaggle competitions with principles that
remain valuable despite its age (in ML time ;)). I saw an
issue on the upstream repo concerning headers, so I
decided to share the version I had in the case it helps others.
It deserved to have outlines/headers (much easier to quickly navigate).
- If you also want the printed version, you can support Abhishek by buying the physical copy. Check the official links in the main/upstream repo here
In case it wasn't clear, I'm not the author of the book, refer to the original repo: https://github.com/abhishekkrthakur/approachingalmost if you have issues concerning the book or all the relevant infos.
Ignore my aaamlp folder, it's just my code following along and isn't a substitute. It's missing datasets, isn't entirely 1:1 with the pdf, I'm not using Tensorflow at all, and my venv is different with missing or different library versions. Fork the upstream/original repo instead!