Fast Stochastic Kalman Gradient Descent for Reinforcement Learning.
Simone TotaroAnders JonssonPublished in: L4DC (2021)
Keyphrases
- reinforcement learning
- direct policy search
- learning automata
- cost function
- stochastic approximation
- function approximation
- state space
- filtering algorithm
- model free
- optimal policy
- markov decision processes
- learning algorithm
- continuous state spaces
- stochastic optimization
- reinforcement learning algorithms
- markov decision process
- linear quadratic
- objective function
- learning process
- kalman filter
- temporal difference
- stochastic processes
- dynamic programming
- multi agent
- action space
- temporal difference learning
- machine learning
- control system
- transition model
- robotic control
- monte carlo