Table 1 The values of R2 in different NN models and indicators are compared.

From: Combining multi-objective genetic algorithm and neural network dynamically for the complex optimization problems in physics

Training sets and model

Number of trainable parameters

R2

\({f}_{\mathrm{FM}}\)

\({R/Q}_{\mathrm{FM}}\)

\({R}_{\mathrm{a FM}}\)

\({\mathrm{Q}}_{\mathrm{FM}}\)

\({f}_{\mathrm{HOM}}\)

3000 samples

Transformer with ANN

265

0.987

0.985

0.989

0.988

0.984

2-layers ANN

76

0.923

0.954

0.957

0.969

0.942