Faster saddle-point optimization for solving large-scale Markov decision processes.
Joan Bas-SerranoGergely NeuPublished in: CoRR (2019)
Keyphrases
- markov decision processes
- saddle point
- transition matrices
- variational inequalities
- policy iteration
- primal dual
- penalty function
- state space
- optimal policy
- dynamic programming
- reinforcement learning
- decision theoretic planning
- stochastic shortest path
- maximum margin
- markov decision problems
- average reward
- linear programming problems
- discrete space
- structured prediction
- interior point
- numerical methods
- dirichlet distribution
- sensitivity analysis
- partially observable
- constrained optimization
- solving problems
- multistage
- least squares
- quadratic programming
- global constraints