Solving Deep Reinforcement Learning Benchmarks with Linear Policy Networks.
Annie WongJacob de NobelThomas BäckAske PlaatAnna V. KononovaPublished in: CoRR (2024)
Keyphrases
- reinforcement learning
- markov decision problems
- optimal policy
- sequential decision making
- function approximators
- policy search
- action selection
- function approximation
- partially observable
- markov decision process
- state and action spaces
- markov decision processes
- quadratic programming
- convex quadratic programming
- reinforcement learning problems
- inverse problems in image processing
- partially observable environments
- actor critic
- action space
- state space
- reinforcement learning agents
- algebraic equations
- policy gradient
- control policy
- policy iteration
- learning algorithm
- policy evaluation
- approximate dynamic programming
- agent learns
- combinatorial optimization
- sufficient conditions
- dynamic programming
- search heuristics
- inverse reinforcement learning
- transition model
- average reward
- linear systems
- least squares
- social networks