A little book named Searching and Ranking, I have composed for several years, esp. during 2011 - 2016. It has been uploaded on horsehour.com. I expected the book could be evolved such that it can be very helpful for those new learners of machine learning. Because of my own interest, all kinds of things are concentrated on Ranking. As for the Searching part, it comes from the search engines, where lost of ranking algorithms have been developed and greatly affect people's online lives.
It's a collection of related research fields. I will continue to refine the book, and provide an online version such that more people could benefit from it. Plus, it can also provide a way to communicate with people from different backgrounds.
It contains about 9 chapters, and each chapter serves a role to introduce a different perspective of solving the ranking problem. Several of them cover the topics of machine learning, mathematical optimization and statistics. These are crucial to understand the content, and also be very important knowledge for new comers to get started as soon as possible. Most people do not like math, but it's the very tool that guide you to better understand the high level ideas. I do not intend to scare you, and where you start the journey depends on your own choice.
The original text are all compiled PDFs from tex files. I like to take notes using LaTex, until I find another very convenient editor: Typora, where I can fluently write text without considering the structural issues. Therefore, I choose to use markdown to reorganize all the documents. However, my only concern is the rendering of mathematical formula. Markdown does not do very well until now. I will try my best to overcome such limitation and produce a higher quality manuscript.
The related repositories you may interested in are irank, metasearch, ivote, socialchoice, and iextract.
Please enjoy it :-)