Development and application of Few-shot learning methods in materials science under data scarcity

Abstract

Machine learning, as a significant branch of artificial intelligence, shortens the cycle of material discovery and synthesis by exploring the characteristics of data.