| Popis: |
Obesity and overweight are major risk factors for a variety of chronic diseases, including cardiovascular diseases like heart disease and stroke, which are the main leading causes of most deaths worldwide. Obesity can also lead to diabetes and its complications, such as blindness, limb amputations, and the need for dialysis. Diabetes prevalence has quadrupled worldwide since 1980. Excess weight can also cause musculoskeletal disorders such as osteoarthritis. The objective of this research is to analyze and predict obesity using machine learning algorithms to assist clinicians and public health agents to make an optical decision related to the prevention, the detection, and the treatment of obesity. Five machine learning classification models including Random Forest, Support Vector Machine, Logistic regression, K-nearest Neighbor, and Ridge Classifier were used for the purpose. These five models were trained after the Exploratory Data Analysis and the Data Preprocessing with k-ford crossvalidation, classification report, the confusion matrix, and the learning curve as metrics. After the training according to the accuracy performance given by each model and the learning curve, the Support Vector Machine was selected and optimized as the final model with 97% of accuracy. |
