Milo Spencer-Harper

simple-neural-network. A neural network written in Python, consisting of a single neuron that uses back propagation to learn.

neural-network-animation. Watch a neural network think. Written in Python.

multi-layer-neural-network. A multi layer neural network written in Python, which can be trained to solve the XOR problem.

visualise-neural-network. Generates a static diagram of a neural network, where each neuron is connected to every neuron in the previous layer.

self-writing-code. Code which randomly mutates in a virtual environment of 100x100 squares with a random distribution of food. If the code doesn't eat, it dies. The code which lives the longest is saved.

neural-network-show-absolute-error. A demonstration of how you can graph the error of the neural network decreasing over time.

neural-network-with-bias. Demonstration of bias in a neural network.

deep-q-learning. A demonstration of how Deep Q Learning can solve the NChain-v0 game.