Prediction of magnetic remanence of NdFeB magnets by using novel machine learning intelligence approach — Support vector regression

A novel model using support vector regression (SVR) combined with particle swarm optimization (PSO) integrating leave-one-out cross validation (LOOCV) was employed to construct mathematical model for prediction of the magnetic remanence of the NdFeB magnets. The leave-one-out cross validation of SVR model test results show that the mean absolute error doesnot exceed 0.0036, the mean absolute percentage error is only 0.53%, and the correlation coefficient (R²) is as high as 0.839. This investigation suggests that the SVR-LOOCV is not only an effective and practical method to simulate the remanence of NdFeB, but also a powerful tool to optimize designing or controlling the experimental process.