A Machine Learning Approach to Predict the Trend of Obesity Prevalence at a Global Level

Excessive weight is associated with adverse health risks. Understanding the global trends in Obesity, in children and adolescents, is paramount to impede the increasing rate of global Obesity prevalence. Different machine learning models were used to predict Obesity prevalence, in children and adolescents, at a global level. This paper presents a novel approach to predict Obesity beyond 2030 using machine learning. The data was derived from a global population-based survey in 2015. In the main study, we applied machine learning models to predict the exponential rise in Obesity prevalence across the world in 2030, 2040, and 2050. In the second study, we further calculated the Obesity prevalence rates according to Socio-Demographic Index (SDI). We obtained promising results with model prediction accuracies of up to 99% R2 for the main study, and up to 92% R2 for the SDI study.