Regret Bounds for Stochastic Shortest Path Problems with Linear Function Approximation.
Daniel VialAdvait ParulekarSanjay ShakkottaiR. SrikantPublished in: ICML (2022)
Keyphrases
- function approximation
- regret bounds
- shortest path problem
- temporal difference learning algorithms
- reinforcement learning
- online learning
- linear regression
- lower bound
- shortest path
- function approximators
- learning tasks
- temporal difference
- upper bound
- combinatorial optimization problems
- multiple objectives
- radial basis function
- model free
- bregman divergences
- optimal solution
- directed acyclic graph
- e learning