Matlab project for fairspace
MATLAB
Reinforcement Learning toolbox
Simevents
This project was written and tested in Matlab2019a however, it should run in the later versions as well.
To train the network edit parameters given below in train_ddpg.m or train_dqn.m and run
max_episodes=2000;
sample_time=0.1;
stopTime=40;
interval=1;
continous_action=false;
Sample time is how fast the simulink model is run (except for the Sim Events parts). Interval is how quickly the reinforcement learning agent is running.