Combination of Machine Learning Techniques to Predict Overweight/Obesity in Adults

Abstract

(1) Background: Artificial intelligence using machine learning techniques may help us to predict and prevent obesity. The aim was to design an interpretable prediction algorithm for overweight/obesity risk based on a combination of different machine learning techniques. (2) Methods: 38 variables related to sociodemographic, lifestyle, and health aspects from 1179 residents in Madrid were collected and used to train predictive models. Accuracy, precision, and recall metrics were tested and compared between nine classical machine learning techniques and the predictive model based on a combination of those classical machine learning techniques. Statistical validation was performed. The shapely additive explanation technique was used to identify the variables with the greatest impact on weight gain. (3) Results: Cascade classifier model combining gradient boosting, random forest, and logistic regression models showed the best predictive results for overweight/obesity compared to all machine learning techniques tested, reaching an accuracy of 79%, precision of 84%, and recall of 89% for predictions for weight gain. Age, sex, academic level, profession, smoking habits, wine consumption, and Mediterranean diet adherence had the highest impact on predicting obesity. (4) Conclusions: A combination of machine learning techniques showed a significant improvement in accuracy to predict risk of overweight/obesity than machine learning techniques separately.

Keywords: artificial intelligence; machine learning; overweight/obesity; predictive model.

Figure 1

Cascade classifier operation flow. Basis structure of proposed cascade classifier, combining gradient boosting (classifier 1), random forest (classifier 2), and logistic regression (classifier 3) machine learning techniques.

Figure 2

Bayesian ACC and density ACC using instances. Probability of wining for each classical machine learning technique separately and the cascade model. Panel (A): Results represented as Bayesian average coverage criterion. Panel (B): Results represented as density average coverage criterion.

Figure 2

Bayesian ACC and density ACC using instances. Probability of wining for each classical machine learning technique separately and the cascade model. Panel (A): Results represented as Bayesian average coverage criterion. Panel (B): Results represented as density average coverage criterion.

Figure 3

SHAP values. The model´s interpretation. Panel (A): Feature importance plot (after feature normalization). Panel (B): SHAP summary plots of the cascade flow model. Each row in the SHAP summary plot represents a feature, with the corresponding Shap values displayed along the x-axis. The features are ranked according to their average absolute Shap values, which represent the most important features of the model. Each point in the plot corresponds to a sample, with the color indicating the magnitude of the feature value, where red denotes larger values and blue denotes smaller values. Panel (C): Waterfall plot for explaining an individual´s prediction results in the validation cohort. The y axis shows the name of the variables and the x axis shows the Shap value. The red bar shows the positive contribution of the feature to the predicted value, and the blue bar shows the negative contribution of the feature to the predicted value.

Figure 3

SHAP values. The model´s interpretation. Panel (A): Feature importance plot (after feature normalization). Panel (B): SHAP summary plots of the cascade flow model. Each row in the SHAP summary plot represents a feature, with the corresponding Shap values displayed along the x-axis. The features are ranked according to their average absolute Shap values, which represent the most important features of the model. Each point in the plot corresponds to a sample, with the color indicating the magnitude of the feature value, where red denotes larger values and blue denotes smaller values. Panel (C): Waterfall plot for explaining an individual´s prediction results in the validation cohort. The y axis shows the name of the variables and the x axis shows the Shap value. The red bar shows the positive contribution of the feature to the predicted value, and the blue bar shows the negative contribution of the feature to the predicted value.