ABSTRACT
With the advancement of technology, data acquisition speed and storage capacity have increased significantly which opens up the unprecedented possibility of real-time course corrections and an accelerated exploration of high-dimensional design spaces, rapid optimization of properties, and a deeper understanding of fundamental science through data-driven approaches and high-throughput experimentation. This data-driven revolution and, in particular, Bayesian-inspired material discovery and synthesis optimization, are poised to transform materials science and technologies evolving from it. To expedite the acceleration of materials discovery the traditional process started with existing experimental databases. Human efforts to extract data from more reliable material datasets and clean the experimental data, or machine learning-assisted material discovery from failed data become an interesting endeavor. The closed-loop materials discovery and synthesis optimization using the Bayesian approach is the future.
