TomBolton.io

Tom Bolton’s AI and Machine Learning Lab Notebook.

checkers

Machine Learning

So, with my new softmax back propagation sorted, I was excited to see how my model would perform, especially since softmax with MSE was my first hypothesis. Well, MSE was a disaster, although I didn’t push it too far. Softmax with…

Machine Learning

Note: This post is the ultimate result of my quest for how to do softmax backpropagation in my hand-coded model for this project. The actual math for softmax back propagation is not something that was specifically covered in my coursework….

Machine Learning

…and it is not lost on me now that I may have had a problem with my gradients which I did not check… –  Me, five days ago After taking a brief hiatus from my checkers AI to do some…

Machine Learning

Having established in my previous post that softmax looks like the way to go for my final activation layer it’s time to think about the cost function. And this one is trickier. Hypothesis: Use Mean Squared Error Cost Function The…

Machine Learning

Back in this post I alluded to the fact that I hadn’t yet written out the approach I’m going to use to calculate costs not just for a single output as in Andrej Karpathy’s Pong from Pixels but from 48 outputs representing…

Machine Learning

This morning, I completed the “plumbing” for checkers. Specifically, with all the game mechanics in place, I now have a framework that will allow two “players” (a red player and a black player) to play against each other ad infinitum…

Machine Learning

There’s a gap  in the courses I can take right now, so I have about two weeks to push forward with checkers and I started again on that yesterday. As I was starting to put together some of the classes…

Machine LearningVideo Games

Last night, after working some more on my checkers algorithm, before going to bed I decided to unwind by playing a little Destiny 2. With checkers and Destiny 2 on my brain, I had a dream in which I was…

Machine Learning

Andrej Karpathy’s Pong from Pixels has been my exemplar for implementing reinforcement learning. I find that in general, I’m following along with his approach, and I feel that conceptually, it’s making sense. But as I set out to actually create my…

Machine Learning

So with these classifiers and unambiguous accuracy numbers when compared to ground truth, it’s dead simple to evaluate success. But what about this checkers game? Unlike the pong from pixels example, my game won’t be playing against a functioning, trained…