Adolescent BMI Prediction using an XGBoost and Artificial Neural Network Hybrid Model
Year: 2026
Type: conference-paper
Abstract: The aim of this research is accuracy improvement, to compare the XGBOOST + ANN Hybrid performance with the Support Vector Deep HealthNet which Predictive model provides best accuracy, stability, and reliability for predictive analysis. Material Methods: In this study, two predictive patterns are evaluated for adolescent BMI prediction. Group 1 uses the existing Deep Health Net model, which applies traditional techniques to classify BMI based on health data. Group 2 uses the proposed XG Boost + Artificial Neural Network (ANN) Hybrid model to predict BMI using adolescent health and lifestyle data collected from 10,000 samples, including Body Mass Index lifestyle attributes, and about health parameters. Result of the XGBoost algorithm improved the average accuracy increased from 88% to 96%, giving an 8% better range compared to the Support deep HealthNet It shows better learning ability and stable prediction for this study. Conclusion: Overall, the XGBoost+ANN Hybrid method provides better predictive performance based on the result from the samples. The suggested XGBoost +ANN Hybrid Predictive model can have a better efficiency than the Support deep HealthNet.
Language: English
Subfield: Health Information Management
Field: Health Professions
Domain: Health Sciences
Open Access status: hybrid
OpenAlex ID: W7203489401