Reinforcement learning in optimization of complex processes

In this talk, I will summarize recent experiments with application of reinforcement learning algorithms, like PPO and DQN, to adaptive optimization of 2 complex processes: cancer treatment and traffic signal control. In both cases, the experiments were carried out in simulation environments properly calibrated using real-world data. I will also outline plans for future development of these techniques, including their integration with metaheuristics and surrogate optimization.

----------------------------------------

We do not know yet if the seminars in forthcoming spring term will be held in the classroom or on-line only at https://meet.google.com/jbj-tdsr-aop

Nie zapadła jescze decyzja czy seminaria w semestzre letnim będą się odbywać stacjonarnie czy wyłącznie online pod adresem https://meet.google.com/jbj-tdsr-aop

---------------------------------------

Plan wystąpień w tej edycji jest dostępny tutaj

The schedule of presentations can be checked here