🧠 Hyperparameter Search Compared: Grid vs Random vs Bayesian Optimization (2D)
Top-down contour map of a validation-loss landscape with grid, random and Bayesian search trials, a Gaussian-process mean or expected-improvement map and regret curves over 40 landscapes.
⚙ Under the hood
Top-down contour map of a validation-loss landscape with grid, random and Bayesian search trials, a Gaussian-process mean or expected-improvement map and regret curves averaged over 40 landscapes. Shares its model core with the 3D page.
hyperparameter tuninggrid searchrandom searchBayesian optimizationregretmachine learning
2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install