# Data‐Driven Fatigue Prediction of Superalloys: A Novel Strategy Integrating Transfer Learning and Partial Label Learning for Addressing Ambiguous Data

## TL;DR

A new machine learning method improves fatigue prediction in superalloys by handling ambiguous data and revealing material relationships.

## Contribution

Integrates partial label learning and transfer learning with thermodynamic calculations for ambiguous data in materials science.

## Key findings

- The method achieves superior predictive accuracy for superalloy fatigue performance.
- Enriched microstructural features improve model interpretability and generalization.
- The framework shows robustness across experimental validation and potential for broader materials applications.

## Abstract

Machine learning has emerged as a powerful tool for predicting material properties due to its efficiency and accuracy. However, challenges related to data integrity, particularly the presence of ambiguous data, have limited its broad application. In this work, a novel strategy is proposed that integrates partial label learning and transfer learning to accurately address ambiguous compositional data in predicting the fatigue performance of superalloys. Subsequently, key microstructural features are enriched through thermodynamic calculations based on the composition data, enhancing model interpretability by revealing composition‐microstructure‐property relationships. This approach not only achieves superior predictive accuracy but also exhibits robust generalization across experimental validation. Given the widespread presence of ambiguous data, this framework holds significant potential for broader applications in materials science.

A novel machine learning strategy tackles ambiguous compositional data to predict superalloy fatigue. By integrating partial label with transfer learning and enriching features through thermodynamic calculations, this approach achieves superior accuracy and interpretability. This robust, generalizable framework pioneers a new path for materials discovery, effectively addressing the prevalent challenge of data ambiguity in advanced materials research.