The organisation of this course revolves around machine learning-related algorithmic problems. Modern machine learning systems are frequently built upon unprovable algorithms, and it is debatable as to when and why they function. This course focuses on creating algorithms whose performance can be thoroughly examined for basic machine learning issues.
The “Algorithmic Aspects of Machine Learning” course, which was taught at MIT in Fall 2013, Spring 2015, and Fall 2017, served as the basis for the book.
This open book has an Open Publication License attached to it (OPL). Algorithmic Aspects of Machine Learning is a free PDF ebook that you can download (1.5 MB).