Teknomo's Q Learning Tutorial MATLAB Code is adapted for our coursework. Graphs are added and Eligibility Traces is incorporated
File 1 ReinformcentLearningGreedy.m
This file has normal Q Learning with Epsilon Greedy Policy The csv file being read in matlab code has to be changed depending on Case (Reward Matrix) The values of constant Epsilon , Gamma, Alpha to be changed as per Case and which states to be read if punishment incorpoated in Case 5A and 5B in code line 49
File 2 ReinforcementLearningQlambdaWatkins.m
It is Watkin's Q Learning Algorithm implementation
Reward Matrixes in CSV Files RewardMatrix25 was just for testing and can be ignored
By: Muaaz Bin Sarfaraz and Chadi El-Hajj These files should be refered with the actual coursework report.
email: [email protected]
Ref:http://people.revoledu.com/kardi/tutorial/ReinforcementLearning/
MAZE to be solved by agent
