Skip to yearly menu bar Skip to main content


Poster
in
Workshop: AI4Mat-2024: NeurIPS 2024 Workshop on AI for Accelerated Materials Design

Known Unknowns: Out-of-Distribution Property Prediction in Materials

Nofit Segal · Aviv Netanyahu · Rafael Gomez-Bombarelli · Pulkit Agrawal

Keywords: [ extrapolation ] [ Property prediction ]

[ ] [ Project Page ]
 
presentation: AI4Mat-2024: NeurIPS 2024 Workshop on AI for Accelerated Materials Design
Sat 14 Dec 8:15 a.m. PST — 5:20 p.m. PST

Abstract:

Developing high-performance materials often requires identifying materials with property values that lie outside the known distribution. Therefore, the ability to extrapolate to out-of-support material property values is invaluable to materials design. Given chemical compositions and their property values, our objective is to learn a predictor that extrapolates zero-shot to higher ranges. In this work, we employ a transductive approach to property prediction and explore its extrapolation capabilities. We leverage analogical composition-target relations in the training and test sets, enabling generalization beyond the training target support.

Chat is not available.