Table 2 Summary of machine learning datasets (A, B, C) constructed in this study

From: Machine learning enabled accurate prediction of structural and magnetic properties of cobalt ferrite

Dataset

Configurations

Variable descriptor

Target property

A

62 structures with different cation distribution

23-component vector:

number of M1-O-M2 bond

DFT-predicted system energies

B

150 magnetic ordering configurations

46-component vector:

number of M1-O-M2 with two different spin alignments

DFT-predicted atomic magnetic moments

C

The same 150 configurations as in B

Product of cation magnetic moments of M1-O-M2 bond

DFT-predicted magnetic stabilization energies

  1. Dataset A is used to predict equilibrium cation distribution as a function of temperature, Dataset B is used to predict magnetic moments based on the local chemical environment, and Dataset C is used to determine superexchange constants of CoFe2O4 crystal.
Back to article page