Fig. 1: Schematic representation comparing black-box and gray-box Bayesian optimization. | npj Computational Materials

Fig. 1: Schematic representation comparing black-box and gray-box Bayesian optimization.

From: A physics informed bayesian optimization approach for material design: application to NiTi shape memory alloys

Fig. 1

In the gray-box approach, the Bayesian optimization framework is enriched with underlying physical information derived from within the system under study. By integrating statistical data with physical insights, a more data-efficient design framework is achieved. Consequently, the optimum design is uncovered with fewer experimental iterations, and the overall modeling uncertainty is significantly reduced.

Back to article page