Technology Category
Machine Learning: Ensemble MethodsData Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal DataKnowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
Enhancing Bagging Ensemble Regression with Data Integration for Time Series-Based Diabetes Prediction
🤖 AI Summary
This study addresses the challenge of accurate temporal forecasting of diabetes prevalence at the U.S. city level, tackling real-world issues including missing and inconsistent multi-source heterogeneous data and sparse ground-truth labels. We construct a standardized, comprehensive diabetes feature dataset covering major U.S. cities from 2011 to 2021. To this end, we propose EBMBag+, the first time-aware enhanced Bagging ensemble model: it unifies heterogeneous data sources via data-driven feature engineering and introduces time-weighted resampling alongside dynamic base-learner fusion—enhancing temporal generalization while preserving ensemble robustness. Evaluated on real-world data, EBMBag+ achieves MAE = 0.41, RMSE = 0.53, MAPE = 4.01%, and R² = 0.90, outperforming six baselines—including SVM, BDTree, LSBoost, NN, LSTM, and the benchmark ERMBag. The model delivers an interpretable, deployable predictive tool for precision public health interventions.
Technology Category
Application Category
📝 Abstract
Diabetes is a chronic metabolic disease characterized by elevated blood glucose levels, leading to complications like heart disease, kidney failure, and nerve damage. Accurate state-level predictions are vital for effective healthcare planning and targeted interventions, but in many cases, data for necessary analyses are incomplete. This study begins with a data engineering process to integrate diabetes-related datasets from 2011 to 2021 to create a comprehensive feature set. We then introduce an enhanced bagging ensemble regression model (EBMBag+) for time series forecasting to predict diabetes prevalence across U.S. cities. Several baseline models, including SVMReg, BDTree, LSBoost, NN, LSTM, and ERMBag, were evaluated for comparison with our EBMBag+ algorithm. The experimental results demonstrate that EBMBag+ achieved the best performance, with an MAE of 0.41, RMSE of 0.53, MAPE of 4.01, and an R2 of 0.9.
Problem
Research questions and friction points this paper is trying to address.
Predicting diabetes prevalence using time series data
Integrating incomplete diabetes datasets for comprehensive analysis
Improving accuracy with enhanced bagging ensemble regression model
Innovation
Methods, ideas, or system contributions that make the work stand out.
Data integration for comprehensive diabetes feature set
Enhanced bagging ensemble regression for time series
Superior performance with low error metrics
🔎 Similar Papers
No similar papers found.
