Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2017-04-17T21:35:20Z
No. of bitstreams: 1
AntoninoAlvesFeitosaNeto_TESE.pdf: 1592913 bytes, checksum: abef6d747d95842e55a8f4f5a5d73859 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2017-04-18T22:17:06Z (GMT) No. of bitstreams: 1
AntoninoAlvesFeitosaNeto_TESE.pdf: 1592913 bytes, checksum: abef6d747d95842e55a8f4f5a5d73859 (MD5) / Made available in DSpace on 2017-04-18T22:17:06Z (GMT). No. of bitstreams: 1
AntoninoAlvesFeitosaNeto_TESE.pdf: 1592913 bytes, checksum: abef6d747d95842e55a8f4f5a5d73859 (MD5)
Previous issue date: 2016-12-09 / Este trabalho aborda a constru??o de comit?s de classifica??o atrav?s t?cnicas metaheur?sticas
de otimiza??o tradicionais de h?bridas. O problema de classifica??o de padr?es ?
tratado como um problema de otimiza??o procurando encontrar o subconjunto de atributos
e classificadores do problema que minimize o erro de classifica??o do comit?. Os comit?s
s?o gerados a partir da combina??o das t?cnicas de k-NN, ?rvore de Decis?o e Naive Bayes
utilizando o voto majorit?rio. Os atributos dos classificadores base s?o modificados pelas
metaheur?sticas de algoritmos gen?ticos, algoritmos mem?ticos, PSO, ACO, M?ltiplos
Rein?cios, GRASP, Simulated Annealing, Busca Tabu, ILS e VNS. Tamb?m s?o aplicados
algoritmos provenientes da arquiteturas de metaheur?sticas h?bridas AMHM e MAGMA.
S?o desenvolvidos algoritmos dessas metaheur?sticas nas vers?es mono e multi-objetivo.
S?o realizados experimentos em diferentes cen?rios mono e multiobjetivo otimizando o erro
de classifica??o e as medidas de boa e m? diversidade. O objetivo ? verificar se adicionar
as medidas de diversidade como objetivos de otimiza??o resulta em comit?s mais acurados.
Assim, a contribui??o desse trabalho ? determinar se as medidas de boa e m? diversidade
podem ser utilizadas em t?cnicas de otimiza??o mono e multiobjetivo como objetivos
de otimiza??o para constru??o de comit?s de classificadores mais acurados que aqueles
constru?dos pelo mesmo processo, por?m utilizando somente a acur?cia de classifica??o
como objetivo de otimiza??o. Verificamos que as metaheur?sticas desenvolvidas apresentam
melhores resultados que as t?cnicas cl?ssicas de gera??o de comit?s, isto ?, Bagging,
Boosting e Sele??o Rand?mica. Verificamos tamb?m que na maioria das metaheur?sticas o
uso das medidas de diversidade como objetivos de otimiza??o n?o auxilia na gera??o de
comit?s mais acurados que quando utilizado somente o erro de classifica??o como objetivo
de otimiza??o obtendo nos melhores cen?rios resultados n?o estatisticamente diferentes. / This work deals with the construction of classification committees using traditional and
hybrid meta-heuristics of optimization techniques. The problem of pattenr classification is
treated as an optimization problem, trying to find the subset of attributes and classifiers
of the problem that minimizes the classification error of the committee. Committees are
generated by combining the techniques of k-NN, Decision Tree and Naive Bayes using the
majority vote. The attributes of the base classifiers are modified by the metaheuristics
of genetic algorithms, memetic algorithms, PSO, ACO, Multi Start, GRASP, Simulated
Annealing, Tabu Search, ILS and VNS. We also apply algorithms from AMHM and
MAGMA hybrid metaheuristics architectures. Algorithms of these metaheuristics are
developed in mono and multi-objective versions. Experiments are performed in different
mono and multiobjective scenarios optimizing classification error and measures of good and
bad diversity. The goal is to verify that adding diversity measures as optimization goals
results in more accurate committees. Thus, the contribution of this work is to determine
if the measures of good and bad diversity can be used in mono and multiobjective
optimization techniques as objectives of optimization for the construction of committees
of classifiers more accurate than those constructed by the same process, but using only the
accuracy classification as an optimization objective. We have verified that the developed
metaheuristics present better results than the classic generation techniques of committees,
ie, Bagging, Boosting and Random Selection. We also verified that in the majority of
metaheuristics the use of diversity measures as optimization objectives does not help to
generate more accurate committees than when only the classification error, obtaining in
the best scenarios non statistically different results.
| Identifer | oai:union.ndltd.org:IBICT/oai:repositorio.ufrn.br:123456789/22684 |
| Date | 09 December 2016 |
| Creators | Feitosa Neto, Antonino Alves |
| Contributors | 66487099449, http://lattes.cnpq.br/1357887401899097, Carvalho, Andr? Carlos Ponce de Leon Ferreira de, 45841888404, http://lattes.cnpq.br/9674541381385819, Ara?jo, Daniel Sabino Amorim de, 04634747405, http://lattes.cnpq.br/4744754780165354, Gouvea, Elizabeth Ferreira, 81652011749, http://lattes.cnpq.br/2888641121265608, Santos, Eulanda Miranda dos, 58017968272, http://lattes.cnpq.br/3054990742969890, Canuto, Anne Magaly de Paula |
| Publisher | PROGRAMA DE P?S-GRADUA??O EM SISTEMAS E COMPUTA??O, UFRN, Brasil |
| Source Sets | IBICT Brazilian ETDs |
| Language | Portuguese |
| Detected Language | English |
| Type | info:eu-repo/semantics/publishedVersion, info:eu-repo/semantics/doctoralThesis |
| Source | reponame:Repositório Institucional da UFRN, instname:Universidade Federal do Rio Grande do Norte, instacron:UFRN |
| Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0032 seconds