On Boundedness of Q-Learning Iterates for Stochastic Shortest Path Problems.
Huizhen YuDimitri P. BertsekasPublished in: Math. Oper. Res. (2013)
Keyphrases
- shortest path problem
- stochastic approximation
- shortest path
- single source
- reinforcement learning
- combinatorial optimization problems
- interval data
- multiple objectives
- cooperative
- state space
- multi agent
- function approximation
- directed graph
- monte carlo
- learning algorithm
- directed acyclic graph
- continuous state spaces
- action selection
- learning automata
- learning rate
- potential field
- dynamic programming
- neural network
- machine learning