UCB-driven Utility Function Search for Multi-objective Reinforcement Learning.
Yucheng ShiAlexandros AgapitosDavid LynchGiorgio CruciataCengis HasanHao WangYayu YaoAleksandar MilenovicPublished in: CoRR (2024)
Keyphrases
- utility function
- multi objective
- reinforcement learning
- decision problems
- evolutionary algorithm
- risk aversion
- multi attribute
- utility maximization
- expected utility
- optimization algorithm
- search space
- multi objective optimization
- optimization criterion
- preference elicitation
- decision theory
- social welfare
- set of pareto optimal solutions
- probability distribution
- decision makers
- multiple objectives
- data mining
- risk averse
- np hard
- quasiconvex
- utility elicitation
- markov decision processes