You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A modeling capability we should add are functions that have stochastic outputs, but not necessarily stochastic inputs (e.g., a Gaussian process). A motivating example pertains to subproblems in Bayesian optimization approaches that embed learned Gaussian processes that serve as stochastic surrogate models.
The text was updated successfully, but these errors were encountered:
A modeling capability we should add are functions that have stochastic outputs, but not necessarily stochastic inputs (e.g., a Gaussian process). A motivating example pertains to subproblems in Bayesian optimization approaches that embed learned Gaussian processes that serve as stochastic surrogate models.
The text was updated successfully, but these errors were encountered: