Average cost temporal-difference learning
The authors propose a variant of temporal-difference learning that approximates average and differential costs of an irreducible aperiodic Markov chain. Approximations are performed through linear combinations of fixed basis functions whose weights are incrementally updated during a single endless trajectory of the Markov chain. In addition, a proof of convergence and a characterization of the limit of convergence are presented. A bound on the resulting approximation error that exhibits an interesting dependence on ``mixing time of the Markov chain is provided.
- On average versus discounted reward temporal-difference learning
- Learning algorithms for Markov decision processes with average cost
- Reinforcement learning based algorithms for average cost Markov decision processes
- Reinforcement learning for long-run average cost.
- Model-based average reward reinforcement learning
- Kernel-based reinforcement learning in average-cost problems
- Differential Temporal Difference Learning
- Average reward reinforcement learning: foundations, algorithms, and empirical results
- Linear least-squares algorithms for temporal difference learning
- Linear least-squares algorithms for temporal difference learning
- Projected equation methods for approximate solution of large linear systems
- Natural actor-critic algorithms
- On average versus discounted reward temporal-difference learning
- A time aggregation approach to Markov decision processes
- Fundamental design principles for reinforcement learning algorithms
- Finite-sample analysis of nonlinear stochastic approximation with applications in reinforcement learning
- A stability criterion for two timescale stochastic approximation schemes
- Reinforcement learning based algorithms for average cost Markov decision processes
- Approximate policy iteration: a survey and some new methods
- Multiscale Q-learning with linear function approximation
- Hyperbolically Discounted Temporal Difference Learning
- An analysis of temporal-difference learning with function approximation
- An online actor-critic algorithm with function approximation for constrained Markov decision processes
- Long-Term Reward Prediction in TD Models of the Dopamine System
- Finite-time performance of distributed temporal-difference learning with linear function approximation
- Scalable Reinforcement Learning for Multiagent Networked Systems
- Risk-Sensitive Reinforcement Learning via Policy Gradient Search
- Efficient multi-objective reinforcement learning via multiple-gradient descent with iteratively discovered weight-vector sets
- Derivatives of logarithmic stationary distributions for policy gradient reinforcement learning
- Actor-critic algorithms with online feature adaptation
- Adaptive data-aware utility-based scheduling in resource-constrained systems
- Stochastic approximation and reinforcement learning: the interface and a little beyond
This page was built for publication: Average cost temporal-difference learning
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1805802)
