PPO agent with continuous action example

I go through the example using PPO to build land rocket model. (https://www.mathworks.com/help/reinforcement-learning/ug/train-ppo-agent-to-land-rocket.html?s_tid=blogs_rc_4) However, the action in this example is discrete. I wonder when I change my action to continuous action . How do I create actornetwork as there is numact as one parameter in actor_network

Answer 1

[Emmanouil Tzorakoleftherakis]

Hello,

If you want to use PPO, i.e. a stochastic actor with continuous action space, you can follow the structure shown here.