Locally Constrained Policy Optimization for Online Reinforcement Learning in Non-Stationary Input-Driven Environments.
Pouya HamadanianArash Nasr-EsfahanySiddartha SenMalte SchwarzkopfMohammad AlizadehPublished in: CoRR (2023)
Keyphrases
- non stationary
- reinforcement learning
- optimal policy
- policy search
- adaptive algorithms
- markov decision process
- concave convex procedure
- action selection
- partially observable
- empirical mode decomposition
- online learning
- dynamic programming
- autoregressive
- state space
- learning algorithm
- function approximators
- partially observable environments
- markov decision processes
- optimal control
- reinforcement learning problems
- markov decision problems
- white noise
- function approximation
- reinforcement learning algorithms
- model free
- exploration exploitation tradeoff
- financial time series
- control policy
- finite horizon
- stock price