This is the project for the MSBD-5012 course. In this project, we will train an agent by the techniques of Reinforcement Learning (Q-Learning) to find the optimal path to the HKUST Logo in the maze.

For further demonstration, please refer to the Youtube Link: HKUSTurkey

The following picture shows the overview of this project:

1. Firstly, use the file envirnment.yml to set up a conda environment called hkust-env. To be more detail, you can go to the directory and execute the following command in your terminal.

conda env create -f environment.yml

2. Secondly, after the installation, execute the command line source activate hkust-env（Mac/Linux）or activate hkust-env（Windows）to activate the environment so that you can run the source code.