Why a textbook?
May of my classes are centered around using models to understand brain and cognition. Sadly, there exists no textbook that I could use. Some common textbooks (Sutton and Barto for reinforcement learning; Abbot and Dayan for neural networks) are expensive and will be used only for very specific pieces of my class. So, most of the time, I have used a collection of review papers to cover the different topics, which turned out to be quite unsatisfactory. A couple of years back, I bit the bullet and wrote my own free “textbook”. It is still a work-in-progress, but here it is.
Why is it online?
It is online, rather than printed, for a variety of reasons. First, it is not polished. Far from it. In fact, I appreciate any feedback on any of its aspects. Second, it is not only made of printed material. All the figures are generated by actual, running demo code that is available on Jupyter Notebooks. And, finally, I want to keep updating it often.
Why is it free?
Well, even if it was a printed textbook, I would not be able to profit from it (legally, a professor cannot profit from a textbooks assigned to a their students). So, I saw no point in even trying to publish it.
The entire book can be downloaded here: Exploratory Models in Cognitive Neuroscience 
Contents
1. Introduction
HTML Text, PDF, Jupyter Notebook