Artificial Intelligence-Assisted Experimental Optimization of Water Oxidation Catalysts

Abstract

Artificial intelligence (AI) methods are very often used to make predictions for datasets that were created externally in arbitrary experiments or on already literature known datasets. In this work, we try to make use of active learning techniques to search for an optimal strategy for the startup-phase of bulk nickel electrodes in the oxygen evolution reaction. The data collected was afterwards reduced in dimensions and used to extract additional information that were learned via an artificial neural network (ANN) on the dataset, respectively.