Adaptive: parallel active learning of mathematical functions

Abstract

Large scale computer simulations are time-consuming to run and often require sweeps over input parameters to obtain a qualitative understanding of the simulation output. These sweeps of parameters can potentially make the simulations prohibitively expensive. Therefore, when evaluating a function numerically, it is advantageous to sample it more densely in the interesting regions (called adaptive sampling) instead of evaluating it on a manually-defined homogeneous grid. Such adaptive algorithms exist within the machine learning field. These methods can suggest a new point to calculate based on all existing data at that time; however, this is an expensive operation. An alternative is to use local algorithms—in contrast to the previously mentioned global algorithms—which can suggest a new point, based only on the data in the immediate vicinity of a new point. This approach works well, even when using hundreds of computers simultaneously because the point suggestion algorithm is cheap (fast) to evaluate. We provide a reference implementation in Python and show its performance.

Publication
Preprint