Accelerated discovery of near-zero ablation ultra-high temperature ceramics via GAN-enhanced directionally constrained active learning

In materials science, a significant correlation often exists between material input parameters and their corresponding performance attributes. Nevertheless, the inherent challenges associated with small data obscure these statistical correlations, impeding machine learning models from effectively capturing the underlying patterns, thereby hampering efficient optimization of material properties. This work presents a novel active learning framework that integrates generative adversarial networks (GAN) with a directionally constrained expected absolute improvement (EAI) acquisition function to accelerate the discovery of ultra-high temperature ceramics (UHTCs) using small data. The framework employs GAN for data augmentation, symbolic regression for feature weight derivation, and a self-developed EAI function that incorporates input feature importance weighting to quantify bidirectional deviations from zero ablation rate. Through only two iterations, this framework successfully identified the optimal composition of HfB 2 -3.52SiC-5.23TaSi 2 , which exhibits robust near-zero ablation rates under plasma ablation at 2500 ​°C for 200 ​s, demonstrating superior sampling efficiency compared to conventional active learning approaches. Microstructural analysis reveals that the exceptional performance stems from the formation of a highly viscous HfO 2 -SiO 2 -Ta 2 O 5 -HfSiO 4 -Hf 3 (BO 3 ) 4 oxide layer, which provides effective oxygen barrier protection. This work demonstrates an efficient and universal approach for rapid materials discovery using small data. In this work, a novel active learning framework combining GAN-based data augmentation and symbolic regression, with a developed EAI acquisition function, successfully identified an optimal UHTC formulation (HfB 2 -3.52SiC-5.23TaSi 2 ) in just two iterations. This composition demonstrated robust near-zero ablation rates at 2500 ​°C, highlighting the superior sampling efficiency of the developed framework over traditional active learning. • GAN-enhanced active learning enables rapid discovery of near-zero ablation UHTCs. • Novel EAI function with directional constraints optimizes HfB 2 -based formulations. • Framework shows broad applicability for materials discovery with small data.