Using Machine Learning to Determine Obesity-Related Risk Factors

Nearly 30% of global mortality will be attributed to lifestyle-related diseases by 2030, according to the World Health Organization (WHO). However, the potential consequences of this outcome can be mitigated through the skillful identification of correlated risk factors and the implementation of behavioral intervention strategies. The prioritization of health behavior modifications is of utmost importance in mitigating the likelihood of life-threatening injuries. The primary objective is not to propose a risk prediction model; instead, it is to conduct a comprehensive analysis of various machine learning (ML) approaches and their potential applications with publicly accessible health data samples. Lifestyle diseases, including obesity, cardiovascular diseases (CVDs), and type II diabetes, will be the primary focus. The study’s participants are aged between 20 and 60 years and of both sexes; individuals who are expectant or who are influenced by genetic factors are excluded. The objective was to employ data science methodologies in order to evaluate the interrelated risk factors associated with obesity and overweight. From “Kaggle” and the “University of California, Irvine (UCI) database,” were obtained and analysed. Furthermore, we conduct an examination of the responses of these potential risk variables to alterations in body-energy imbalances through the utilisation of regression analysis and data visualisation methodologies. No novel risk factors were identified through the evaluation of pre-existing data on obesity and overweight using machine learning algorithms. However, it contributed to our comprehension of the relationships between identified risk factors and weight fluctuations, in addition to the methods for graphically representing these associations. The efficacy of classification and regression models is improved when they are implemented on a reduced subset of the dataset.