Search Results

Now showing 1 - 4 of 4

Some of the metrics are blocked by your
consent settings
Prototype selection using reinforcement learning and minimal consistent subset identification guide
(2010-01-01)
Kruatrachue, Boontee
;
Choowong, Teeratorn
This paper try to apply Reinforcement Learning (RL) to a task with large number of states. This usually is a difficult task since RL has less chance to visit all state or has enough number of visit to learn average reward accurately. Moreover, RL may not be able to learn or obtain any optimal solution as RL learn by averaging rewards from each action performing in each state. In order to alleviate this RL learning problem, any solution to a task such as, non-optimal algorithm or heuristics can collaborate with RL by using their knowledge to prune the non-optimal action in each state. This reduces search space of RL and helps it learn faster. A Minimal consistent subset problem is used as an example to demonstrate how RL can learn faster with the help of other heuristics. ©ICROS.
Some of the metrics are blocked by your
consent settings
Episode continuation and exploration start for reinforcement learning with large number of states
(2009-01-01)
Kruatrachue, Boontee
;
Anantapornkit, Ekaphol
This paper presents simple techniques to improve the learning rate of the RL algorithm in a task with large number of states that the algorithm cannot fully explore by going through all possible states. The problem with the standard RL algorithm is that varying the ε parameter in the ε-greedy policy is not sufficient to improve its learning performance. The techniques allow the algorithm to focus on exploring the local path to eventually obtain more rewards and occasionally switching to another path to avoid being trapped in the local region of the search space. In order to investigate the effectiveness of the techniques, the minimal consistent subset identification (MCSI) problem is used as a test problem. The paper concludes by comparing the size of the identified subset obtained from the standard RL algorithm and the proposed algorithm along with those of the standard MCSI method.
Some of the metrics are blocked by your
consent settings
Prototype selection based on minimal consistent subset and genetic algorithms
(2008-12-01)
Kruatrachue, Boontee
;
Hongsamart, Marut
This paper applies the genetic algorithms to identify the minimal "consistent" prototype subset [1]. This subset can be used as a prototype which correctly recognizes the entire original prototype set. This proposed genetic algorithm tries to And the minimal consistent subset to reduce recognition time in nearest neighbor [2] classification. The main difference from other genetic algorithm (GA) approaches is the hybrid of minimal consistent set identification (MCSI) method [3] and genetic algorithm. The MCSI method provides the local optimal number of prototype while the Genetic performs the global search. The proposed hybrid algorithm has been tested on several problems and compared with the results of MCSI and other GA approach [4]. © 2008 SICE.
Some of the metrics are blocked by your
consent settings
Minimal consistent subset selection as integer nonlinear programming problem
(2006-12-01)
Kangkan, Kamonnat
;
Kruatrachue, Boontee
The minimal consistent subset selection is a solution of high computational demands problem of the nearest neighbor decision system. This paper presents a new approach that aims to make the problem more clearly by stating it as a constrained optimization problem, called "integer nonlinear programming problem (INLP)". In this context, we propose method that formulates the minimal consistent subset selection problem as 0-1 integer nonlinear programming problem. We show experimental result of the minimal consistent subset of "IRIS Dataset", obtained by solving its constrained optimization model. The results obtained suggest that the approach offers exactly optimal solution of the problem. © 2006 IEEE.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results