# Property-aimed embedding: a machine learning framework for material discovery

## TL;DR

This paper introduces a machine learning framework that enhances material discovery by enabling property-specific, quantitative predictions, especially useful when data is abundant for general knowledge but scarce for specific samples.

## Contribution

The proposed property-aimed embedding framework advances material discovery by integrating property-specific learning with limited sample scenarios, outperforming traditional empirical criteria.

## Key findings

- Achieves comparable differentiation of alloy phases with simple data-driven criteria.
- Effective in scenarios with abundant general data but scarce specific samples.
- Flexible and applicable to various material discovery tasks.

## Abstract

Proposing new materials by atom substitution based on periodic table similarity is a conventional strategy of searching for materials with desired property. We introduce a machine learning frame work that promotes this paradigm to be property-specific and quantitative. It is of peculiar usefulness in situations where abundance data is accessible for learning general knowledge but samples for the problem of interest are relatively scarce. We showcase its usage and viability in the problem of separating high entropy alloys with different structural phases, for which a very simple data-driven criterion achieves differentiating ability comparable with widely used empirical criteria. Its flexibility and generability make it a promising tool in other material discovery tasks and far beyond.