Analyzing patient data...
Predict Diabetes Risk
3
Clinical Overview
Real-time diabetes risk intelligence — May 2026
● Live
⚠ Educational model only — XGBoost trained on PIMA Indians dataset. Not for clinical diagnosis.
📊
Total Predictions
0
↑ 12% this month
⚠
High Risk Cases
0
↑ 8% vs last week
✓
Model AUC-ROC
0.82
XGBoost tuned
👥
Dataset Size
768
PIMA dataset
Prediction Trend Last 7 days
Trend data.
Risk Distribution All patients
42% high risk.
Age Distribution
Age groups.
Model Performance
Key metrics — tuned XGBoost
Accuracy
74%
AUC-ROC
82%
Recall
78%
Precision
60%
F1 Score
68%
Risk Prediction
Sliders and fields are fully synced — changes update the live risk estimate instantly
● XGBoost Model
Live Risk Estimate
35%
MEDIUM
Low RiskMedium RiskHigh Risk
Glucose
120
BMI
30.0
Age
33
Pedigree
0.47
Patient Data Input
Pregnancies
?
0 – 17
Glucose (mg/dL)
?
40 – 250
Blood Pressure (mm Hg)
?
40 – 140
Skin Thickness (mm)
?
1 – 99
Insulin (μU/mL)
?
1 – 846
BMI
?
10 – 70
Diabetes Pedigree Function
?
0 – 2.42
Age (years)
?
18 – 100
Prediction Result
—
—%
Diabetes Probability
Prediction
—
Confidence
—
AI Clinical Insights
⚕Educational model only. Results must be reviewed by a qualified clinician before any medical decision.
⚕
Ready for Analysis
Adjust sliders or type values on the left.
The live bar updates instantly. Hit
Analyze Patient Risk for the full result.
Dataset Analytics
PIMA Indians Diabetes Dataset — 768 records
Class Distribution
268 diabetic.
Feature Importance
Glucose highest.
Glucose vs BMI Risk Scatter
Risk stratification by two key features
Risk clusters visible.
Patient Registry
Recent predictions and outcomes
Patient Age Glucose BMI Risk Score Level Date
Model Performance
Experiment comparison — XGBoost variants
Experiment Results
Model Comparison (AUC-ROC)
SMOTE highest.
Confusion Matrix — Tuned XGBoost
Settings
Platform configuration
API Configuration
Status: waiting for test...
Profile