Classificação do desempenho do atleta de judô utilizando regressão logística e rede neural

Detalhes bibliográficos
Ano de defesa: 2019
Autor(a) principal: Teixeira, Felipe Guimarães
Orientador(a): Não Informado pela instituição
Banca de defesa: Não Informado pela instituição
Tipo de documento: Tese
Tipo de acesso: Acesso aberto
Idioma: por
Instituição de defesa: Universidade Federal do Rio de Janeiro
Brasil
Instituto Alberto Luiz Coimbra de Pós-Graduação e Pesquisa de Engenharia
Programa de Pós-Graduação em Engenharia Biomédica
UFRJ
Programa de Pós-Graduação: Não Informado pela instituição
Departamento: Não Informado pela instituição
País: Não Informado pela instituição
Palavras-chave em Português:
Link de acesso: http://hdl.handle.net/11422/13239
Resumo: The present study aims to classify judo athletes in national and regional levels using logistic regression (LR) and multilayer perceptron neural network (MLP) on anthropometric and biomechanical data, as well as the special judo fitness test (SJFT). 42 competitive two level judo male athletes (28 national and 14 state level) were submitted to the following measures and tests: (a) skinfold thickness; (b) circumferences; (c) bone widths; (d) stabilometric test; (e) SJFT; and (f) dynamometry. The RL and MLP models were used to classify the two levels of judo athletes. Before adjusting the models, the forward stepwise method was used to select the variables that produced the highest performance. To further reduce the number of variables, a combinatorial analysis was performed over the variables previously selected. The RL and MLP models presented area under ROC curve 91.0% to 96.0% and 82% to 89%, respectively. The three variables that best classified the groups were epicondylar humerus width, total number of throws on the SJFT, and stabilometric mean velocity of center of pressure in mediolateral direction. This study demonstrated that LR and MLP could be used to classify judo athletes from national and state competitive levels, using a reduced set of anthropometric, biomechanical, and SJFT variables.