Risk-sensitive MPCs with Deep Distributional Inverse RL for Autonomous Driving.
Keuntaek LeeDavid IseleEvangelos A. TheodorouSangjae BaePublished in: IROS (2022)
Keyphrases
- risk sensitive
- autonomous driving
- model free
- markov decision processes
- optimal control
- reinforcement learning
- grand challenge
- stereo vision
- control policies
- markov decision problems
- reinforcement learning algorithms
- utility function
- vision algorithms
- policy iteration
- optimal policy
- state space
- action space
- temporal difference
- urban traffic
- learning algorithm
- function approximation
- control strategies
- control strategy
- average reward
- optimality criterion
- average cost
- decision theoretic
- dynamic programming
- optimal solution
- decision making
- decision processes
- probability distribution
- multi agent systems