- This Q-Learning code for MATLAB has been written by Ioannis Makris and Andrew Chalikiopoulos. It trains an agent to find the shortest way through a 25x25 maze. Following convergence of the algorithm, MATLAB will print out the shortest path to the goal and will also create three graphs to measure the performance of the agent.reinforcement-learning-robot-in-maze-master.zip
- Reinforcement learning, a Q learning algorithm, implementation on a robot that tryies to solve randomly created maze and reach the goal. Note that you can run .m files both on Matlab and Octave.Q-Learning-master.zip
- Successfully implemented Q-Learning for a simple robot navigation problem of a robot moving on a 5 x 5 grid with one arbitrary goal (reward of +10) and three arbitrary obstacles (reward of -10)deep_q_rl-master.zip
- This package provides a Lasagne/Theano-based implementation of the deep Q-learning algorithm described in:
Playing Atari with Deep Reinforcement Learning Volodymyr Mnih, Koray Kavukcuoglu, David Silver, Alex Graves, Ioannis Antonoglou, Daan Wierstra, Martin Riedmiller
Mnih, Volodymyr, et al. "Human-level control through deep reinforcement learning." Nature 518.7540 (2015): 529-533.
Here is a video showing a trained network playing breakout (using an earlier version of the code):