Method and computer for experimental design

摘要

A method for experimental design comprises: A) selection of at least one first experiment from an experiment range by a data-driven optimizer; B) input of data for the first experiment that has been determined by experimentation; C) use of a meta-layer for evaluating the experiment data; D) input of the experiment data determined by experimentation for the first experiment into the data-driven optimizer; E) influencing of the data-driven optimizer by the result of the evaluation; F) selection of at least one second experiment from the experiment range by the data-driven optimizer. The meta-layer contains a neuronal network and/or a hybrid model and/or a rigorous model and/or a data mining method.