GAUSSIAN PROCESS + BAYESIAN OPTIMIZATION
Gaussian Process Surrogate Optimizer Lab
Fit a visual Gaussian process surrogate from sparse x:y observations, inspect posterior uncertainty, compute expected improvement, and export the recommended next experiment with SVG, CSV, Markdown notes and JSON receipt.
Reviewed 2026-06-28
MathRBF kernelPosterior mean95% uncertaintyExpected improvementCSV exportsJSON receipt
FORMULA CONTRACT
A useful optimizer shows uncertainty, not only a winning point.
The lab uses an RBF kernel k(x,x') = sigma_f^2 exp(-(x-x')^2 / 2l^2), Gaussian process posterior equations and expected improvement EI(x). The recommended next x is the grid point with the highest expected improvement.