Aggrandizing the accuracy of body fat percentage by stratification using decision tree

Abstract

Bariatrics is the branch of science which deals with obesity and its related surgical procedures. A person’s physical indolence, unhealthy food habits and genetic constitution emanates as the fons et origo of health gremlins. Multifaceted indagations have been worked on the diverse and heterogenous obstinate concerns caused due to obesity. Anatomization of body fat percentage has become a rudimentary regimen for every individual to be done in a fastidious manner. The whilom work anent body fat percentage entailed Body Mass Index (BMI) with respect to age and gender of a person. The anatomical conformation of an individual unraveling the fat constitution and the muscle tissue composition is not computationally enumerated using BMI. Thus, the formula using BMI dempers the veracity for a person having more muscle mass than fat mass and speciously vitiates the fat percentage of that person. The proposed novel formula is analyzed by cross-validated classification model using decision tree, and is effectuated by implementing information gain. This accentuates the coherence, efficacy and accuracy of the derived body fat percentage for a person. The Ethical Committee approval for this study has been obtained from the Institutional Ethics Committee, Madras Medical College, Chennai. The empirical study has been simulated using Matlab and the results have been successfully acquired in the GUI mode.