23 search results - page 2 / 5 » Stochastic Local Search for POMDP Controllers |
FROCOS
2011
Springer
12 years 4 months ago
2011
Springer
A dominant approach to Satisfiability Modulo Theories (SMT) relies on the integration of a Conflict-Driven-Clause-Learning (CDCL) SAT solver and of a decision procedure able to h...
AAAI
2000
13 years 6 months ago
2000
Reinforcement learning (RL) can be impractical for many high dimensional problems because of the computational cost of doing stochastic search in large state spaces. We propose a ...
NIPS
2003
13 years 6 months ago
2003
We describe a new approximation algorithm for solving partially observable MDPs. Our bounded policy iteration approach searches through the space of bounded-size, stochastic fini...
JAIR
2006
13 years 5 months ago
2006
In this paper, we introduce DLS-MC, a new stochastic local search algorithm for the maximum clique problem. DLS-MC alternates between phases of iterative improvement, during which...
HICSS
1994
IEEE
13 years 9 months ago
1994
IEEE
We are developing a new problem-solving methodology based on a self-organization paradigm. To realize our future goal of self-organizing computational systems, we have to study co...