Application of Machine Learning in Amorphous Alloys.

Area of Science:

• Materials Science
• Computational Materials Science
• Alloy Design

Background:

• Traditional amorphous alloy development relies on empirical trial-and-error and density functional theory (DFT), facing limitations in efficiency and development time.
• Modern research demands faster, more cost-effective methods for exploring amorphous alloy systems and predicting material properties.

Purpose of the Study:

• To review the key applications of machine learning (ML) in the field of amorphous alloys.
• To highlight ML's advantages in accelerating the design, analysis, and property prediction of amorphous alloys.

Main Methods:

• Review of existing literature on machine learning applications in amorphous alloys.
• Focus on four primary ML applications: amorphous alloy phase prediction, amorphous composite phase prediction, glass-forming ability (GFA) prediction, and material property prediction.

Main Results:

• Machine learning demonstrates significant advantages over traditional methods, including lower experimental costs, enhanced performance, and reduced development cycles.
• ML is effectively applied to predict amorphous alloy phases, composite phases, GFA, and material properties.

Conclusions:

• Machine learning is a transformative tool for amorphous alloy research, offering superior efficiency and cost-effectiveness.
• Future directions include developing advanced ML models, integrating ML with high-throughput experimentation, and establishing standardized data-sharing platforms for materials science.