Estimation of Obesity Levels

Abstract — In this report, we estimate obesity levels in individuals from the countries of Mexico, Peru, and Colombia, based on their eating habits and physical condition. Descriptive analysis, correlation analysis, statistical tests, and regression analysis have been conducted using the dataset. Logistic regression has been used to perform a regression analysis to classify the obesity level of an individual. It is a multi-class classification problem where the dataset we use derives from the Obesity Level dataset. We created a logistic regression model that classifies an individual into one of seven obesity target values with a macro average f1 score of 0.71.