Abstract: Overcoming the trade-off between saturation magnetic induction (B_s) and coercivity (H_c) of Fe-based nanocrystalline alloys (FNAs) remains a great challenge due to the traditional design relying on trial-and-error methods, which are time-consuming and inefficient. Herein, we present an interpretable machine learning (ML) algorithm for the effective design of advanced FNAs with improved B_s and low H_c. Firstly, the FNAs datasets were established, consisting of 20 features including chemical composition, process parameters, and theoretically calculated parameters. Subsequently, a three-step feature selection was used to screen the key features that affect the B_s and H_c of FNAs. Among six different ML algorithms, extreme gradient boosting (XGBoost) performed the best in predicting B_s and H_c. We further revealed the association of key features with B_s and H_c through linear regression and SHAP analysis. The valence electron concentration without Fe, Ni, and Co elements (VEC1) and valence electron concentration (VEC) ranked as the most important features for predicting B_s and H_c, respectively. VEC1 had a positive impact on B_s when VEC1 < 0.78, while VEC had a negative effect on H_c when VEC < 7.12. Optimized designed FNAs were successfully prepared, and the prediction errors for B_s and H_c are lower than 2.3 % and 18 %, respectively, when comparing the predicted and experimental results. These results demonstrate that this ML approach is interpretable and feasible for the design of advanced FNAs with high B_s and low H_c.

Key words: Nanocrystalline alloy, Machine learning, Feature selection, Saturation magnetic induction, Coercivity