Classifier hierarchy learning by means of genetic algorithms

Martínez-Otzeta, José María; Sierra, Basilio; Lazkano, Elena; Astigarraga, Aitzol

doi:10.1016/j.patrec.2006.06.001

Cited by 20 publications

(13 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this paper we continue the work presented in [8], where the authors define a new hierarchy construction method and make a search in the space of all possible hierarchies by means of a genetic algorithm. Here we show that good results could also be achieved by a pure random search.…”

Section: Introductionmentioning

confidence: 91%

Analyzing Classifier Hierarchy Multiclassifier Learning

Martínez-Otzeta

Sierra

Lazkano

et al. 2008

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Abstract. Classifier combination falls in the so called machine learning area. Its aim is to combine some classification paradigms in order to improve the individual accuracy of the component classifiers. Classifier hierarchies are an alternative among the several methods of classifier combination. In this paper we present new results about a recently proposed hierarchy construction method. Experiments have been carried out over 42 databases from the UCI repository, showing an improvement over the performance of the base classifiers.

show abstract

Section: Introductionmentioning

confidence: 91%

Analyzing Classifier Hierarchy Multiclassifier Learning

Martínez-Otzeta

Sierra

Lazkano

et al. 2008

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…From this came our first idea to apply a genetic algorithm to a clustering problem, while trying to find the best solution. This idea was especially supported first, when comparing genetic algorithm clustering results with those obtained by hill-climbing search methods [67][68][69][70][71] and second, when reviewing the literature concerning MEDLINE abstract clustering and finding that a genetic algorithm had never been applied. Since genetic algorithms do not usually fit with textual data structures, data preprocessing is necessary.…”

Section: Proposed Algorithmmentioning

confidence: 99%

MEDLINE Text Mining: An Enhancement Genetic Algorithm Based Approach for Document Clustering

Karâa

Ashour

Sassi³

et al. 2015

Intelligent Systems Reference Library

View full text Add to dashboard Cite

MEDLINE is the largest biomedical literature database. It is updated daily with 200-4,000 citations. This permanent growth induces the need of a good MEDLINE abstract clustering to accelerate the procedure of research and information retrieval. Several works have been developed in this context, but clustering MEDLINE abstracts are still an area where researchers are trying to propose new approaches to better clustering. Over the last few years, evolutionary algorithms have been widely applied to clustering problems because of their ability to avoid local optimal solutions and converge to a global one. In this article, a new approach is proposed for clustering MEDLINE abstracts based on an extension of an evolutionary algorithm which is the genetic algorithm combined with a Vector Space Model and an agglomerative algorithm.

show abstract

“…Furthermore, there are some studies about using GA to build decision trees that are provided accordingly. Classifier hierarchy learning by means of GAs [49], optimizing prediction models by GA (based on decision trees and neural networks) [50], optimization of decision tree classifier through GA [51], fitness of binary decision tree by GA [52], classification tree analysis using TARGET [1], and utilization of the elitist multi-objective genetic algorithm for classification rule generation [53].…”

Section: Literature Reviewmentioning

confidence: 99%

“…Using classifier (in this case: decision tree) hierarchies. It is an alternative method among several methods to combine classifiers [49]. This hierarchy is used to arrange single classifiers in a tree.…”

Section: Parametermentioning

confidence: 99%