Projecting and Comparing Obesity Trends Worldwide via Machine Learning Models

Obesity, recognized as a chronic disease, impacts a significant segment of the global population, with about 40% of U.S. adults affected, and similar trends observed in diverse ethnic communities. Internationally, nations like Mexico and Saudi Arabia experience rising obesity rates, largely attributable to shifts in dietary habits, while Japan and South Korea report lower prevalence, reflecting differing lifestyle choices. The research posits that early detection and intervention through predictive diagnostics could revolutionize obesity management. The current study, through the use of IBM Watson’s analytics tools, aims to examine Latin American obesity data to compare significant factors to worldwide trends. The focus on identifying key differences that contribute to obesity, such as environmental and lifestyle factors, is crucial to enhance understanding and improve intervention strategies. Logistic regression, Random Forest, and Snap Logistic classifier algorithms from the IBM Watson platform aided in evaluating data from 2,112 patients across Mexico, Columbia, and Peru, by analyzing variables such as age, gender, lifestyle habits, and dietary intake, demonstrating high accuracies between 90.9% and 96.9, in predicting obesity-related outcomes. A comprehensive global dataset was used to compare and contrast the influence of environmental factors on obesity, comparing them with a primary data set. Despite minor improvements with model enhancements, the critical predictors remained largely consistent across different models. This study provides robust evidence that machine learning can effectively contribute to the early detection and potential mitigation of obesity, supporting public health initiatives to address this issue. Expanding on this, the findings scrutinize specific differences and key factors causing obesity based on machine learning predictions. The application of these technologies in real-world settings not only underscores the potential of AI in healthcare but also sets a precedent for future research in the domain of predictive health analytics.