AI Bias Detection Dashboard

Accuracy: ${(metrics.accuracy*100).toFixed(1)}%
Precision: ${(metrics.precision*100).toFixed(1)}%
Recall (TPR): ${(metrics.recall*100).toFixed(1)}%
False Positive Rate: ${(metrics.fpr*100).toFixed(1)}%
Selection Rate: ${(metrics.selectionRate*100).toFixed(1)}%

Instructions:
1. Upload two CSV files: one with the ground truth (correct) data and one with your model's predictions.
2. Both files MUST contain an identical unique ID column so data can be matched.
3. After processing, select your columns and define the "Favorable Outcome" (e.g., 'Hired', '1', 'Loan Approved').

Ground Truth Data (.csv) No file selected.

Model Predictions Data (.csv) No file selected.

AI Bias Detection Dashboard

Configure Data Columns

Overall Performance & Fairness Metrics

Group Analysis

Metric Comparison Across Groups