Enterprise AI Analysis: Multilevel Determinants of Overweight and Obesity Among U.S. Children Aged 10–17: Comparative Evaluation of Statistical and Machine Learning Approaches Using the 2021 National Survey of Children's Health
Unpacking Childhood Obesity: A Comparative AI Study of Multilevel Determinants
This analysis evaluates the predictive power of various machine learning models (logistic regression, random forest, gradient boosting, XGBoost, LightGBM, MLP, TabNet) in identifying multilevel determinants of overweight and obesity among U.S. adolescents (10-17 years) using data from the 2021 National Survey of Children's Health. We found that model discrimination ranged from 0.66 to 0.79, with logistic regression, gradient boosting, and MLP offering the most stable balance of discrimination and calibration. While boosting and deep learning models showed modest improvements in recall and F1 scores, no single model was uniformly superior. Key predictors consistently included behavioral factors (diet, physical activity, sleep), household socioeconomic conditions (parental stress, education, poverty), and neighborhood characteristics. Critically, persistent performance disparities across racial and poverty groups were observed, suggesting that algorithmic complexity alone is insufficient to address underlying data quality issues and structural inequities.
Key Executive Impact & Performance Indicators
0
Highest AUC (Logistic Regression)
0
Models Evaluated
Multilevel
Predictor Domains Identified
Persistent
Subgroup Disparities
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
0.788
Highest AUC Achieved (Logistic Regression)
| Model | AUC | Accuracy | Precision | Recall | F1 | Brier |
|---|---|---|---|---|---|---|
| Logistic Regression | 0.788 | 0.769 | 0.830 | 0.360 | 0.502 | 0.160 |
| Random Forest | 0.763 | 0.765 | 0.822 | 0.350 | 0.491 | 0.166 |
| Gradient Boosting | 0.786 | 0.765 | 0.822 | 0.350 | 0.491 | 0.161 |
| XGBoost | 0.768 | 0.761 | 0.760 | 0.384 | 0.510 | 0.166 |
| LightGBM | 0.772 | 0.768 | 0.793 | 0.384 | 0.518 | 0.164 |
| Deep Learning (MLP) | 0.779 | 0.766 | 0.771 | 0.397 | 0.524 | 0.162 |
| TabNet | 0.655 | 0.736 | 0.767 | 0.264 | 0.393 | 0.194 |
WgtConcn_21
Consistently Strongest Predictor Across Models
| Logistic Regression | LightGBM | TabNet |
|---|---|---|
| WgtConcn_21 (2.894) | WgtConcn_21 (0.117) | WgtConcn_21 (0.111) |
| SC_RACE_R (0.433) | AdultEduc_21 (0.024) | SC_RACE_R (0.082) |
| AdultEduc_21 (0.382) | SC_AGE_YEARS (0.013) | PhysAct_21 (0.052) |
| SC_SEX (0.248) | SC_SEX (0.006) | ScreenTime_21 (0.052) |
| PhysAct_21 (0.181) | SC_RACE_R (0.005) | NbhdAmenities_21 (0.049) |
| povlev4_21 (0.176) | NbhdSupp_21 (0.003) | NbhdDetract_21 (0.047) |
| ParCoping_21 (0.159) | FoodSit_21 (0.002) | AdultEduc_21 (0.046) |
| FoodSit_21 (0.142) | HrsSleep_21 (0.002) | FamCount_21 (0.043) |
| SmkInside_21 (0.127) | ScreenTime_21 (0.002) | povlev4_21 (0.041) |
| FamCount_21 (0.106) | PhysAct_21 (0.001) | ACEincome_21 (0.039) |
Enterprise Process Flow
Raw Dataset
→
Sample Selection
→
Data Pre-processing
→
Missing Data Imputation
→
One-hot Encoding
→
Scale
→
Train-Val-Test Split
→
Variable Construction
→
Train Machine Learning Models
→
Model Evaluation
→
Fairness Evaluation
Addressing Subgroup Disparities in Obesity Prediction
This study reveals persistent disparities in model performance across racial and poverty groups, highlighting a critical limitation in current AI applications for public health. While some deep learning models modestly improved recall for certain disadvantaged subgroups, algorithmic complexity alone did not resolve these inequities. Sex and metropolitan status showed smaller, less consistent differences. The findings underscore that algorithmic advances are insufficient without improvements in data quality, representation, and measurement. Future efforts should prioritize addressing structural inequities and enhancing data quality to achieve true predictive equity and better health outcomes.
Calculate Your Potential AI Impact
Estimate the efficiency gains and cost savings your organization could achieve by implementing tailored AI solutions.
Estimated Annual Savings
$0
Annual Hours Reclaimed
0
Your AI Implementation Roadmap
A typical enterprise AI journey, from initial strategy to scaled impact and continuous improvement.
Phase 1: Discovery & Strategy
In-depth analysis of your current operations, identification of high-impact AI opportunities, and development of a tailored strategic roadmap. This phase includes stakeholder interviews, data assessment, and a comprehensive solution design.
Phase 2: Pilot & Proof of Concept
Development and deployment of a minimal viable AI solution in a controlled environment. Focus on validating core hypotheses, measuring initial performance gains, and gathering user feedback for iterative refinement.
Phase 3: Full-Scale Implementation
Expansion of the AI solution across relevant departments and workflows. Integration with existing enterprise systems, robust data pipeline development, and comprehensive training for end-users and administrators.
Phase 4: Optimization & Scaling
Continuous monitoring of AI model performance, identification of further optimization opportunities, and scaling the solution to maximize enterprise-wide impact. Establishment of governance frameworks for long-term sustainability.
Ready to Transform Your Enterprise with AI?
Schedule a personalized consultation to explore how these insights can be applied to your unique business challenges and opportunities.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat