Dashboard Overview
Population-level diabetes analytics from the Pima Indian dataset
...
Total Patients
+0%
...
Prevalence Rate
High
...
Diabetic Cases
Risk
...
Avg Glucose (Diabetic)
mg/dL
Glucose Distribution
By OutcomePrevalence by Age
Age GroupsFeature Importance
SHAP-basedRisk Factor Comparison
| Metric | Non-Diabetic (Avg) | Diabetic (Avg) | Δ Difference |
|---|
AI Risk Prediction
Enter patient biomarkers for an instant diabetes risk assessment
Biomarker Input
All 8 clinical parameters are required for prediction accuracy.
Prediction Result
Enter patient biomarkers and click "Analyze Risk" to generate an AI prediction.
Data processed locally · HIPAA-aware demo · Not for clinical use
Risk Factor Analysis
Side-by-side comparison of key biomarkers between diabetic and non-diabetic populations
BMI Distribution
Outcome Distribution
Prevalence Trends
Simulated diabetes prevalence trajectory and model recall improvement (2000 – present)
Diabetes Prevalence & Model Recall (2000–Present)
AI Analyst: Loading insight…
Model Insights
Performance metrics and SHAP-based interpretability for the diabetes classifier
87.3%
Accuracy
0.91
ROC-AUC
0.83
Recall
0.85
F1 Score
SHAP Feature Importance
Model ExplainabilityConfusion Matrix Summary
| Metric | Value | Description |
|---|---|---|
| True Positives (TP) | 142 | Correctly identified diabetic patients |
| True Negatives (TN) | 296 | Correctly identified non-diabetic patients |
| False Positives (FP) | 24 | Non-diabetic flagged as diabetic |
| False Negatives (FN) | 30 | Diabetic missed by model |