I want to understand neural networks thoroughly, and help you do so too. Throughout this blog, we will focus on questions of why.
There are lots of tutorials about how to code neural networks, but they often leave behind a trail of why's. Now, that's not to say we'll forgo how's. In fact, we'll build our very own neural network from scratch in Python, much the way Michael Nielsen does in his excellent online book. The difference is that we'll use more graphical illustrations, and what I believe to be friendlier notation. After the basics, we'll dive into coding bleeding-edge papers together, explaining the key insights in terms of concepts we've already covered.
If you have suggestions or comments please hit me up using the links below. Criticism helps me make the best material :)