Structure in the space of value functions

Machine Learning

Volumn 49, Issue 2-3, 2002, Pages 325-346

  Foster, David ^a,b   Dayan, Peter ^b  

^a UNIVERSITY OF EDINBURGH   (United Kingdom)

^b UNIVERSITY COLLEGE LONDON   (United Kingdom)

Author keywords

Density estimation; Dynamic programming; Mixture models; Reinforcement learning; Unsupervised learning; Value functions

Indexed keywords

DENSITY ESTIMATION; MARKOV DECISION PROBLEMS; MIXTURE MODELS; REINFORCEMENT LEARNING; UNSUPERVISED LEARNING; VALUE FUNCTIONS;

DYNAMIC PROGRAMMING; HIERARCHICAL SYSTEMS; MARKOV PROCESSES; MATHEMATICAL MODELS; OPTIMAL CONTROL SYSTEMS;

LEARNING SYSTEMS;

EID: 0036832959     PISSN: 08856125     EISSN: None     Source Type: Journal    
DOI: 10.1023/A:1017944732463     Document Type: Article

References (41)

1. Learning structure of latent variable models by variational Bayes
2. Neuronlike elements that can solve difficult learning problems
3. Learning and sequential decision making
4. Decision theoretic planning: Structural assumptions and computational leverage
5. The Laplacian pyramid as a compact image code
6. Finite element methods for active contour models and balloons for 2d and 3d images
7. Improving elevator performance using reinforcement learning
8. O-Plan: The open planning architecture
9. Feudal reinforcement learning
10. Maximum likelihood from incomplete data via the EM algorithm
11. Hierarchical reinforcement learning with the MAXQ value function decomposition
12. Composing functions to speed up reinforcement learning in a changing world
13. Learning and executing generalized robot plans
14. Multilayer control of large Markov chains
15. Efficient stochastic source coding and an application to a Bayesian network source model
16. Variational inference for Bayesian mixtures of factor analyzers
17. Stable fitted reinforcement learning
18. Planning with temporally abstract actions
19. Hierarchical solution of Markov-decision processes using macro-actions
20. Generative models for discovering sparse distributed representations
21. Autoencoders, minimum description length, and Helmholtz free energy
22. Hierarchical reinforcement learning: Preliminary results
23. Exponentiated gradient versus gradient descent for linear predictors
24. Macro-operators: A weak method for learning
25. Multi-value functions: Efficient automatic action hierarchies for multiple goal MDPs
26. Hierarchical control and learning for Markov decision processes
27. Reinforcement learning with hierarchies of machines
28. Multi-time models for temporally abstract planning
29. Theoretical results on reinforcement learning with temporally abstract options
30. On-line Q-learning using connectionist systems
31. Reinforcement learning with a hierarchy of abstract models
32. Reinforcement learning with soft state aggregation
33. TD models: Modeling the world at a mixture of time scales
34. Between MDPs and semi-MDPs: Learning, planning and representing knowledge at multiple temporal scales
35. Generating project networks
36. Finding structure in reinforcement learning
37. Learning from delayed rewards