Pedestrian simulation as multi-objective reinforcement learning.
Naresh Balaji RavichandranFangkai YangChristopher E. PetersAnders LansnerPawel Andrzej HermanPublished in: IVA (2018)
Keyphrases
- multi objective
- reinforcement learning
- optimization algorithm
- multi objective optimization
- evolutionary algorithm
- multiple objectives
- genetic algorithm
- mathematical model
- machine learning
- robocup soccer
- simulation environment
- function approximation
- multi agent
- particle swarm optimization
- evolutionary optimization
- objective function
- conflicting objectives
- learning algorithm
- multiobjective optimization
- learning process
- pareto optimal
- real robot
- optimum design
- simulation models
- optimal control
- temporal difference
- model free
- pedestrian detection
- pose estimation
- optimal policy
- genetic programming
- state space
- optimal solution