... with Stochastic Gradient Descent.
Neural network stuff: src/nn
Using it on MNIST handwriting number recognition: src/mnist
Success threshold and the learning rate (eta) are defined in src/mnist/index.ts
, play with those if you want.
The most interesting I've found is threshold: 0.05, eta: 2.
This book thing to figure out the SGD related stuff. Converting from gross untyped python code to a language with typing is really annoying.