Somatotype classification using: Neural Networks, Decision Trees, and Logistic Regression
Abstract
Body shape is defined by genetics, diet, and daily exercise. Body shape is important to define and exploit skills in sports. For example, a runner requires an Ectomorphic body, that is, thin and with the least amount of body fat to enhance his speed. On the contrary, a professional wrestler is required to be an Endomorph, which has a lot of fat and muscle. Therefore, classifying body shapes can help identify the ideal areas for each sport. The method for obtaining the somatotype with the Heath-Carter technique is through the measurement of weight, height, circumference of arms, legs, wrists, ankles, among other measurements. With the measurements, calculations are applied to know the somatotype, obtaining parameters of each somatotype. In this work, somatotypes are classified: Ectomorph, Endomorph and Mesomorph. With a Dataset with 618 records of young adults. The Dataset was classified with the Orange tool using an Artificial Neural Network, Decision Trees and Logistic Regression obtaining results of 93% accuracy. It is concluded that it is possible to obtain the classification of somatotypes with the data of the person's measurements without doing the calculations.
Downloads
References
Baldaño, M., & Stanley, S. (marzo de 2011). Somatotipo y deporte, de https://www.efdeportes.com/efd154/somatotipo-y-deporte.htm
López C.I., Domínguez, M., Ávila, L., Galindo, M., & Ching, J. (2015). Antecedentes, descripción y cálculo de somatotipo. Aristas: Investigación Básica y Aplicada, 3(6), 43-49.
Chollet, F. (2018). Deep Learning with Python. Manning. ISBN: 9781617294433.
Drywień M., Górnicky, K., Górnicka M., (2021). Application of Artificial Neural Network to Somatotype Determination. Applied Science, 11(4), 1365. https://doi.org/10.3390/app11041365
Ferreira, V. (19 de 02 de 2021). Treino. (Treino Maestre ) https://treinomestre.com.br/biotipo-e-somatipo-aprenda-o-conceito-correto/
Jahandideh, R, Tarhi A., Tahmasbi M. (2018). Physical attribute prediction using deep residual neural networks. arXiv preprint arXiv. https://doi.org/10.48550/arXiv.1812.07857
Jakšić, D. & Cvetkovic M., (2009). Neural network analysis of somatotype differences among males related to the manifestation of motor abilities. Acta Kinesiologica, 3(1), 107-113.
Jaksic, D., Lilic L., Popovic S., Matic R., Molnar S. (2014). Application of a More Advanced Procedure in Defining Morphological Types. Int. J. Morphol. [online]. 2014, vol.32, n.1, pp.112-118. ISSN 0717-9502. http://dx.doi.org/10.4067/S0717-95022014000100019.
Krzykala, M., Karpowicz, M., Strzelczyk, R., Pluta, B., Podciechowsca, K., & Karpowicz, k. (2020). Morphological asymmetry, sex and dominant somatotype among Polish youth. PLos ONE, 19(9), https://doi.org/10.1371/journal.pone.0238706.
Papandrianos N, Papageorgiou E, Anagnostis A, Papageorgiou K (2020) Bone metastasis classification using whole body images from prostate cancer patients based on convolutional neural networks application. PLoS ONE 15(8): e0237213. https://doi.org/10.1371/journal.pone.0237213
Rivera Bedoya, L. y. (2020). Diseño de un algoritmo de redes neuronales artificiales para la elaboración de planes de acondicionamiento físico personalizados. Universidad del Valle.
Sharma, L., Majumder J. (2013). Application of artificial Neural Network on body somatotype analysis among Indian Population
Copyright (c) 2024 María Fernanda Urdañez Carbajal, José Sergio Ruiz Castilla, Adrián Trueba Espinosa
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
