dc.contributor.advisor |
Lara Cueva, Román Alcides |
|
dc.contributor.author |
Estévez Salazar, Alexis Darío |
|
dc.date.accessioned |
2018-11-13T22:42:12Z |
|
dc.date.available |
2018-11-13T22:42:12Z |
|
dc.date.issued |
2018 |
|
dc.identifier.citation |
Estévez Salazar, Alexis Darío (2018). Towards an automatic detection system of sports talents; an approach to Tae Kwon Do. Carrera de Ingeniería en Electrónica y Telecomunicaciones. Universidad de las Fuerzas Armadas ESPE. Matriz Sangolquí. |
es_ES |
dc.identifier.other |
040533 |
|
dc.identifier.uri |
http://repositorio.espe.edu.ec/handle/21000/15269 |
|
dc.description.abstract |
Tae Kwon Do is a Korean martial art and Olympic combat sport, which is characterized by amazing techniques of kicking. In this sense, it is possible to extract different features of this sport, in this case, we have used well-defined features associates to combat athletes. Herein, we present a support system for national selected athletes team based on feature selection and ranking from Ecuadorian athletes. We use Wrapper and Embedded methods to choose features, which are based on entropy of information and weights of features respectively. For supervised classification, we use two well known algorithms such as Decision Trees and Support Vector Machine. The highest performance was obtained from all features analysis, v-SVM, RBF kernel, v = 0.23 outputs an accuracy of 90.909%, and the key features are Overweight and Technical - tactical abilities. |
es_ES |
dc.language.iso |
eng |
es_ES |
dc.publisher |
Universidad de las Fuerzas Armadas ESPE. Carrera de Ingeniería en Electrónica y Telecomunicaciones. |
es_ES |
dc.rights |
openAccess |
es_ES |
dc.subject |
MACHINE LEARNING |
es_ES |
dc.subject |
WRAPPER - EMBEDDED METHOD |
es_ES |
dc.subject |
PERFORMANCE |
es_ES |
dc.subject |
APRENDIZAJE AUTOMÁTICO |
es_ES |
dc.subject |
MÉTODOS WRAPPER - EMBEDDED |
es_ES |
dc.subject |
RENDIMIENTO |
es_ES |
dc.title |
Towards an automatic detection system of sports talents; an approach to Tae Kwon Do |
es_ES |
dc.type |
article |
es_ES |