58 search results - page 7 / 12 » Fuzzy Approximation for Convergent Model-Based Reinforcement... |
IWLCS
2005
Springer
14 years 1 months ago
2005
Springer
Aiming to clarify the convergence or divergence conditions for Learning Classifier System (LCS), this paper explores: (1) an extreme condition where the reinforcement process of ...
ATAL
2003
Springer
14 years 28 days ago
2003
Springer
Much emphasis in multiagent reinforcement learning (MARL) research is placed on ensuring that MARL algorithms (eventually) converge to desirable equilibria. As in standard reinfor...
ACMICEC
2007
ACM
13 years 11 months ago
2007
ACM
This paper introduces algorithms for learning how to trade using insider (superior) information in Kyle's model of financial markets. Prior results in finance theory relied o...
WSC
2008
13 years 10 months ago
2008
Reinforcement Learning (RL) is a simulation-based technique useful in solving Markov decision processes if their transition probabilities are not easily obtainable or if the probl...
ICML
2000
IEEE
14 years 8 months ago
2000
IEEE
This paper discusses theoretical and experimental aspects of gradient-based approaches to the direct optimization of policy performance in controlled ??? ?s. We introduce ??? ?, a...