Home▸AI & Machine Learning▸Optuna Bayesian Hyperparameter Search Lab

🎯 Optuna Bayesian Hyperparameter Search Lab

Watch a Tree-structured Parzen Estimator explore a 3D loss landscape, comparing TPE Bayesian sampling against random search and pruning of weak trials.

AI & Machine Learning3DAdvanced60 FPS

A synthetic two-parameter loss surface rendered as a 3D purple landscape, where trial markers rain down and settle into valleys as a Tree-structured Parzen Estimator (TPE) — the sampler behind Optuna — learns where to search next.

🔬 What It Demonstrates

Each falling sphere is one Optuna trial. Under TPE, new trials cluster near the best valley found so far instead of scattering uniformly, and weak trials can be pruned early — exactly the mechanics that make Optuna converge faster than grid or random search.

🎮 How to Use

Pick TPE or random search, adjust γ to shift between exploitation and exploration, toggle pruning, and watch the best-loss beam track the study's optimum. Generate a new landscape to see the sampler adapt to a different loss surface.

💡 Did You Know?

Optuna's default sampler models P(params | good) and P(params | bad) separately, then proposes the candidate that maximizes their ratio — a cheap approximation to full Bayesian optimization that scales to dozens of hyperparameters.

⚙ Under the hood

Watch a Tree-structured Parzen Estimator explore a 3D loss landscape, comparing TPE Bayesian sampling against random search and pruning of weak trials.

bayesian optimizationhyperparameter tuningtperandom searchmachine learningoptimizationThree.js

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