公开(公告)号：US20200097853A1

公开(公告)日：2020-03-26

申请号：US16615309

申请日：2017-06-02

Applicant: Google LLC

Inventor： Daniel Reuben Golovin ,  Gregory Peter Kochanski ,  John Karro

IPC: G06N20/00 ,  G06F17/11 ,  G06N7/00

Abstract: A computer-implemented method for black box optimization of parameters of a system, product, or process, can include performing, by one or more computing devices, one or more iterations of a sequence of operations which includes: determining whether to sample an argument value from a feasible set of argument values using a first approach or using a second approach, wherein each argument value of the feasible set defines values for each of plural parameters of a system, product, or process; and based on the determination, sampling the argument value using the first approach or the second approach wherein the first approach includes sampling the argument value at random from the feasible set and the second approach includes sampling, by the one or more computing devices, the argument value from a subset of the feasible set that is defined based on a ball around a current best argument value.