Algorithm 2 Construct training samples

The sampling grid and 8-coil MIT model are established. for number of needed samples do Load the model. for $\left\{{\mathsf{\rho }}^{\left(\mathrm{i}\right)},{\mathsf{\theta }}^{\left(\mathrm{i}\right)}\right\}$ in coordinates of anomalies do for j in coordinates of number of coils (8) anomalies do Assign $\left\{{\mathsf{\rho }}^{\left(\mathrm{i}\right)},{\mathsf{\theta }}^{\left(\mathrm{i}\right)}\right\}$ to change the position of anomaly (anomalies) in the model. The j-th coil acts as the excitation and the others as the receiver. Re geometric modeling and mesh generation. Run the model. 7 voltage data are obtained from the receiving coils. 3409 conductivity data are obtained from the sampling grid. end for end for Vec-operator the collected voltage data and the conductivity data. Save as a corresponding sample $\left\{{\mathrm{V}}^{\left(num\right)},{\mathsf{\sigma }}^{\left(num\right)}\right\}$. end for