Adaptive Risk Sensitive Path Integral for Model Predictive Control via Reinforcement Learning.
Hyung-Jin YoonChuyuan TaoHunmin KimNaira HovakimyanPetros G. VoulgarisPublished in: MED (2023)
Keyphrases
- risk sensitive
- reinforcement learning
- model predictive control
- optimal control
- model free
- markov decision processes
- control system
- function approximation
- utility function
- dynamic programming
- predictive control
- state space
- data mining
- machine learning
- optimality criterion
- markov decision problems
- reinforcement learning algorithms
- control strategy
- infinite horizon
- temporal difference
- control policies
- learning algorithm
- policy iteration
- decision theoretic
- multi agent
- neural network
- optimal policy
- case based reasoning