Home▸AI & Machine Learning▸Bayesian Hyperparameter Search (2D)

🎯 Bayesian Hyperparameter Search (2D)

A 2D top-down heatmap of the same Gaussian-process surrogate: posterior mean colour-mapped across the hyperparameter grid, an uncertainty fog that clears as trials accumulate, and a live acquisition-function view.

AI & Machine Learning2DAdvanced60 FPS📱 Mobile-adapted⇄ 3D version

This 2D companion runs the identical Gaussian-process regression and Upper Confidence Bound acquisition function as the 3D version, but reads it as a top-down heatmap instead of an orbiting surface: posterior mean is colour-mapped from blue (low predicted validation score) through red (high), a dark uncertainty "fog" thickens over any region the model hasn't sampled near and visibly clears as trials land nearby, gold dots mark every trial actually run, and a pulsing indigo ring tracks wherever the acquisition function currently recommends trying next — so the exploration/exploitation trade-off that drives Bayesian hyperparameter search is legible at a glance rather than requiring a camera orbit to read.

⚙ Under the hood

2D top-down companion to the Gaussian-process surrogate: posterior mean and uncertainty rendered as a heatmap with a live fog overlay, the same UCB acquisition function, and a true-landscape comparison view.

bayesian optimizationhyperparameter tuningprobabilitygaussian processsurrogate modelmachine learningcanvas 2d

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