Gaussian process modeling of doped-lanthanum manganites Curie temperature from compositions and ionic radii

Abstract

The response of Curie temperature, TC, upon doping becomes rather complex when both A- and B-site cations are partially substituted in lanthanum manganites. Here, we develop the principal component based Gaussian process regression model to investigate statistical correlations among TC, ionic radii, and compositions. The dataset includes a wide range of single- and multi-doped manganites with TC ranging from 40 K to 375 K. Our model demonstrates a high correlation coefficient, and high prediction accuracy as reflected through the prediction root mean square error and mean absolute error being 0.76% and 0.23% of the average experimental TC, respectively.