Applies the expected improvement transformation to the model output.
Applies the expected improvement transformation to the model output.
Predictive mean output from the model
Predictive variance output from the model
The expected improvement over the current best evaluation
Expected improvement selection criterion. This transformation produces the expected improvement of the model predictions (over the current "best" value).
"Practical Bayesian Optimization of Machine Learning Algorithms" (PBO), https://papers.nips.cc/paper/4522-practical-bayesian-optimization-of-machine-learning-algorithms.pdf