See Dissertation_Final.pdf for an in depth overview of my project.
The aim of my project was to develop a game engine for the card game Castle and an AI player. In order to produce the AI player, I compared and contrasted different algorithms. Upon finding the most appropriate, I then developed a working player before further enhancing it using Deep Learning. After collecting a vast amount of training data, I used Keras to train a Convolutional Neural Network to be used by the AI player to evaluate game states and make better decisions.