Plotting states while doing RL training
2 vues (au cours des 30 derniers jours)
Afficher commentaires plus anciens
Jorn de Vries
le 14 Mar 2023
Réponse apportée : Emmanouil Tzorakoleftherakis
le 14 Mar 2023
Is it possible to plot the states of a state space model vs time, while doing RL training, in this DoubleIIntegrator example? And related to that perhaps, get the time so it can be used as an input for the reward function in the rlEnv?
0 commentaires
Réponse acceptée
Emmanouil Tzorakoleftherakis
le 14 Mar 2023
We recently added a mechanism that allows you to log any information you find helpful during training. Please take a look at this page.
You can use the same mechanism to log time, but with MATLAB environments, unlike Simulink, there is no intrinsic notion of 'time'. You just assume that every time you call step in your environment, 'some' time has passed since the previous call. For some applications it makes sense to consider that these step calls happen at fixed time intervals, for others (e.g. event-driven systems) it may not be the case.
0 commentaires
Plus de réponses (0)
Voir également
Catégories
En savoir plus sur Deep Learning Toolbox dans Help Center et File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!