By Roberto

For this project, I have trained four AI models to play the classic game Snake using a supervised learning method. Two of those models were trained by using neural networks (convolutional neural network and dense network), the other two models were trained by using machine learning algorithms (decision tree and K-nearest neighbors). The objective of this project is to compare these different approaches in terms of performance, by measuring the speed of each model and its high score. All code for this project is written in Python, while using the Tensorflow and Scikit-learn libraries to train the models.

> Read report

Related Posts