Solving Relational MDPs with Exogenous Events and Additive Rewards.
Saket JoshiRoni KhardonPrasad TadepalliAswin RaghavanAlan FernPublished in: ECML/PKDD (1) (2013)
Keyphrases
- markov decision processes
- reinforcement learning
- semi markov decision processes
- transition matrices
- markov decision problems
- relational databases
- relational data
- sequential decision making problems
- reward function
- dynamic programming
- state space
- optimal policy
- decision diagrams
- policy iteration
- algebraic decision diagrams
- finite horizon
- average cost
- finite state
- infinite horizon
- decision theoretic planning
- factored mdps
- stochastic shortest path
- partially observable
- average reward
- multi relational
- multiarmed bandit
- relational learning
- fully observable
- linear programming
- data model
- multi armed bandits
- bayesian networks