gated-SANE: A Constrained Strategic Autonomous Non-Smooth Exploration for multiple optima discovery over Noisy Microscopy Experiments

Both computational and experimental material discovery bring forth the challenge of exploring multidimensional and multimodal parameter spaces, such as phase diagrams of Hamiltonians with multiple interactions, composition spaces of combinatorial libraries, material structure image spaces, and molecular embedding spaces. To address the black-box and time-consuming evaluation, active learning methods such as Bayesian optimization (BO) is a suitable approach. However, the experimental data are often too noisy to initiate a false convergence by the conventional BO where the exploration can get overly focused near a single or fake global optimal, deviating from the broader goal of scientific discovery. To address these limitations, we have developed a Strategic Autonomous Non-Smooth Exploration (SANE) to facilitate an intelligent Bayesian optimized navigation with a proposed cost-driven probabilistic acquisition function to find multiple global and local optimal regions. We expand SANE into gated-SANE via integrating a minor human intervened constraint. We implemented the gated-SANE into different Piezoresponse spectroscopy data which demonstrated better performance than BO to facilitate the exploration and discovery of multiple open and confined optimal regions. Our work showcases the potential application of this method to noisy experiment, where such combined strategic and human intervened approaches can be critical in unlocking new discoveries in autonomous research.