Home▸Machine Learning & Neural Networks▸Group Fairness Lab: ROC Curves per Group and the Impossibility Identity (3D)

⚖️ Group Fairness Lab: ROC Curves per Group and the Impossibility Identity (3D)

3D Blender-built glass columns hold 1,600 synthetic applicants sorted by model score; move per-group thresholds, compare demographic parity, equal opportunity and predictive parity, and check the fairness impossibility identity.

Machine Learning & Neural Networks3DModerate60 FPS📱 Mobile-adapted🟠 Blender⇄ 2D version
Group Fairness Lab: ROC Curves per Group and the Impossibility Identity (3D) ↗ Open standalone
⚙ Under the hood

Blender-built glass columns hold 1,600 synthetic applicants sorted by model score: drag a threshold plate per group, compare demographic parity, equal opportunity and predictive parity on per-group ROC and calibration curves, and see a balance scale tilt with the selection gap. The impossibility identity is checked on closed form and sampled data.

Algorithmic FairnessROC CurveDemographic ParityEqual OpportunityImpossibility Theorem

3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install

What did you find?

Add reproduction steps (optional)