Experience of Use of an Artificial Neural Network as an Auxiliary in the Physical Fitness Classification of Taekwondo Athletes from Mexico City

Evaluation is a continuous process in the activities of an athlete, is part of their training and competition. It serves to know the physical fitness state and to identify skills that can be improved with a specific exercise program. There are few works that relate the results of these tests and their interaction with the efficiency in performing the fundamental motor gesture. In this work, 40 athletes are evaluated through tests of body composition, anaerobic power, balance control and isokinetic muscle strength. The results of the evaluations are grouped in what is proposed as a characteristic vector of the athlete and a new EK index is proposed to evaluate the efficiency of the front kick (Cut). The athlete characteristic vector is used as input values in an artificial neural network (ANN). The EK index is proposed to stratify the output of the ANN, by ordering the athletes and classify them into 4 levels of physical aptitude: perfectible, average, good and optimum. To validate the ANN, a linear regression analysis is used, obtaining a value of R=0.91 and a classification of 100%. This high percentage is due to the homogeneity of the sample as they are high-performance athletes.