Back to Research
Model Card

Model Card

Transparent documentation of what this model does, who it is for, and where it should and should not be used.

Model Details

Model typeXGBoost binary classifier
Version1.0.0
Trained2026
Frameworkscikit-learn + XGBoost
Input features14 BRFSS clinical survey features
OutputBinary (0 = No diabetes, 1 = Diabetes) + probability
ExplainabilitySHAP (TreeExplainer) + LIME (LimeTabularExplainer)

Intended Use

Primary usePopulation-level Type 2 diabetes risk screening
Intended usersHealthcare researchers, public health practitioners, students
Deployment contextResearch and educational demonstration only
Out-of-scope useClinical diagnosis, individual medical decisions without professional oversight

Training Data

SourceCDC Behavioral Risk Factor Surveillance System (BRFSS)
Features14 demographic, lifestyle, socioeconomic, and clinical variables
Target variableSelf-reported diabetes diagnosis (binary)
Class balanceAddressed using SMOTE oversampling on training set
Train/test split80% training, 20% held-out test set (stratified)
Cross-validation5-fold stratified cross-validation

Performance

Accuracy0.7208
F1-Score0.4484
ROC-AUC0.8297
Precision0.3123
Recall0.7943

Limitations

PopulationTraining data over-represents certain demographics (BRFSS survey population)
Self-reported dataFeatures are self-reported, introducing potential recall and reporting bias
TemporalModel reflects disease patterns at time of training — may drift over time
GeographyTrained on US survey data; may not generalise to other health systems or populations
Diagnosis proxyTarget variable is self-reported diagnosis, not laboratory-confirmed

Ethical Considerations

FairnessModel includes race/ethnicity and income as features — SHAP analysis should be monitored for disparate impact across groups
TransparencyAll predictions include SHAP explanations; no black-box outputs to end users
Not a diagnosisThis tool must never be used as a substitute for clinical laboratory testing
Data privacyNo patient data is stored beyond the session unless the user is authenticated and opts in
AccountabilityPredictions should always be reviewed by a qualified healthcare professional before action is taken

⚠ Important Disclaimer

This model is developed for research and educational purposes as part of a Moringa School capstone project. It is not approved for clinical use. Predictions should never replace clinical laboratory testing or professional medical assessment.