Home▸Data Science▸Reliability Diagram Plotter: Confidence vs. Accuracy

Reliability Diagram Plotter: Confidence vs. Accuracy

Interactive 2D reliability-diagram plotter: generate a synthetic classifier's predicted probabilities, dial in over/under-confidence, and watch the confidence-vs-accuracy curve and prediction histogram pull toward the perfect-calibration diagonal as Platt scaling or isotonic regression recalibrate it, with Brier score, ECE and MCE updating live.

Data Science2DAdvanced60 FPS📱 Mobile-adapted⇄ 3D version

A classifier that says "80% confident" should be right about 80% of the time — most aren't. This simulator generates a synthetic classifier's predicted probabilities from a controllable confidence slope, plots a genuine 2D reliability diagram (a confidence-vs-accuracy curve against the perfect-calibration diagonal) together with a jittered strip plot of every raw sample, and lets you apply Platt scaling or isotonic regression to recalibrate it, watching Brier score, Expected Calibration Error and Maximum Calibration Error update live as the curve moves.

⚙ Under the hood

Generate a synthetic classifier's predicted probabilities, dial in over- or under-confidence, and watch Platt scaling and isotonic regression pull a live 2D reliability-diagram curve — plus a jittered per-sample strip plot — onto the perfect-calibration diagonal while Brier score, ECE and MCE update in real time.

calibrationreliability-diagrambrier-scoreplatt-scalingisotonic-regressionmachine-learning

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install