Deep Q-learning

I have currently implemented Q-learning for training to dock a vessel at a given position.

I now want to alter this to be deep Q-learning instead, but can't really understand how to.

Do anyone have code using deep Q-learning for something similar that thet could share so i could have a look at the build up?

Thanks in advance!

Answer 1

[Sean de Wolski]

https://www.mathworks.com/help/releases/R2019a/reinforcement-learning/ug/q-agents.html

[Brage Mothes]

No sorry I still need an example code of deep Q-learning

Answer 2

[Chad MacDonald]

For general information on DQN agents, see Deep Q-Network Agents.

There are several DQN training examples in the documentation:

• Train DQN Agent to Balance Cart-Pole System
• Train DQN Agent to Swing Up and Balance Pendulum
• Train DQN Agent for Lane Keeping Assist

Hopefully, these examples can help get you started.