Hello everyone,
For those who don’t know, I am currently working on a supervised learning/RL learning training session with a 3DOF model of a jet aircraft.
My idea is to change, during a training, some parameters such as the number of allowed interventions of the external controller as well as their duration. Another whish is to change accordingly these parametes as the agent makes progress, e.g. its episode reward will increase during training. Is it possible to do this during a training? That is, is it possible to implement some for loops that exploits the trainStats object while the training is runnig? Thank youHello everyone,
For those who don’t know, I am currently working on a supervised learning/RL learning training session with a 3DOF model of a jet aircraft.
My idea is to change, during a training, some parameters such as the number of allowed interventions of the external controller as well as their duration. Another whish is to change accordingly these parametes as the agent makes progress, e.g. its episode reward will increase during training. Is it possible to do this during a training? That is, is it possible to implement some for loops that exploits the trainStats object while the training is runnig? Thank you Hello everyone,
For those who don’t know, I am currently working on a supervised learning/RL learning training session with a 3DOF model of a jet aircraft.
My idea is to change, during a training, some parameters such as the number of allowed interventions of the external controller as well as their duration. Another whish is to change accordingly these parametes as the agent makes progress, e.g. its episode reward will increase during training. Is it possible to do this during a training? That is, is it possible to implement some for loops that exploits the trainStats object while the training is runnig? Thank you reinforcement learning, supervised learning, machine learning, rl, rl block, trainstats, matlab, simulink MATLAB Answers — New Questions

​