1863 search results - page 2 / 373 » Multiagent learning using a variable learning rate |
ICMAS
1998
13 years 6 months ago
1998
We describe a framework that can be used to model and predict the behavior of MASs with learning agents. It uses a difference equation for calculating the progression of an agent&...
ECAI
2008
Springer
13 years 6 months ago
2008
Springer
In this paper, we propose a policy gradient reinforcement learning algorithm to address transition-independent Dec-POMDPs. This approach aims at implicitly exploiting the locality...
AAAI
2008
13 years 7 months ago
2008
Modern Bayesian Network learning algorithms are timeefficient, scalable and produce high-quality models; these algorithms feature prominently in decision support model development...
ICML
2000
IEEE
14 years 5 months ago
2000
IEEE
This paper presents a general method to derive tight rates of convergence for numerical approximations in optimal control when we consider variable resolution grids. We study the ...
FSS
2006
13 years 4 months ago
2006
Convergence of blind delayed source separation algorithms, which use constant learning rates, is known to be slow. We propose a fuzzy logic based approach to adaptively select the...