Table 5 Details of the results of each model.

Asqu Bpla Cnor Dnsti Ecoa

#43

Train

Successful models: MAML(MSE = 0.49, R² = 0.93) > Fine-tune (MSE = 0.61, R² = 0.92) > MMN (MSE = 0.71, R² = 0.91) > Four-stage model (MSE = 0.76, R² = 0.90) > DNN (all data set (MSE = 2.21, R² = 0.71)) > DNN (target data set (MSE = 3.31, R² = 0.56))

Failed models:none

#65

Train

Successful models: four-stage model (MSE = 0.40, R² = 0.97) > MMN (MSE = 1.19, R² = 0.90) > MAML (MSE = 2.48, R² = 0.79) > DNN(target data set (MSE = 10.391,R² = 0.10))

Failed models: DNN (all data set (MSE = 32.73, R² = 0)), Fine-tune(MSE = 59.82, R² = 0)

#63

Test

Successful models: MMN(MSE = 0.10,R² = 0.99) > Four-stage model(MSE = 1.28,R² = 0.85) > MAML(MSE = 2.87,R² = 0.66) > DNN(target data set (MSE = 3.79, R² = 0.56)) > DNN (all data set(MSE = 3.82, R² = 0.55)) > Fine-tune (MSE = 3.87, R² = 0.55) > 

Failed models: none

#320

Test

Successful models: MMN (MSE = 0.02, R² = 0.59) > MAML (MSE = 0.33, R² = 0.10)

Failed models: four-stage model(MSE = 15.94,R² = 0),DNN (target data set (MSE = 37.84,R² = 0)), DNN (all data set (MSE = 62.56,R² = 0)), Fine-tune (MSE = 94.18,R² = 0)

Asqu Bpla Cnwat Dsti Enor

#77

Train

Successful models: MMN(MSE = 0.40, R² = 0.97) > MAML (MSE = 1.49,R² = 0.88) > Four-stage model (MSE = 1.66, R² = 0.87) > DNN (all data set (MSE = 51.75, R² = 0))

Failed models: DNN (target data set(MSE = 81.85, R² = 0)), Fine-tune (MSE = 91.44, R² = 0)

#78

Train

Successful models: MMN(MSE = 0.19, R² = 0.98) > Four-stage model (MSE = 0.76, R² = 0.94) > MAML(MSE = 4.42,R² = 0.62)

Failed models:DNN (all data set (MSE = 28.17, R² = 0)), DNN (target data set (MSE = 51.84, R² = 0)), Fine-tune (MSE = 67.40, R² = 0)

#73

Test

Successful models:MMN (MSE = 0.29,R² = 0.98) > Four-stage mode l(MSE = 1.27, R² = 0.93)

Failed models: DNN (target data set (MSE = 161.68, R² = 0)), Fine-tune (MSE = 223.61, R² = 0), MAML (MSE = 234.6, R² = 0), DNN (all data set (MSE = 252.98,R² = 0))

#76

Test

Successful models:MMN (MSE = 0.47, R² = 0.97) > Four-stage model (MSE = 1.22, R² = 0.93)

Failed models: DNN (target data set (MSE = 160.70, R² = 0)), Fine-tune (MSE = 222.5, R² = 0), MAML (MSE = 233.4, R² = 0), DNN (all data set (MSE = 251.73, R² = 0))

SRRC task

SRRC#11

Train

Successful models: MMN (MSE = 0.14, R² = 0.98) > DNN (all data set (MSE = 0.80, R² = 0.88)) > Fine-tune (MSE = 1.94, R² = 0.72)

Failed models: four-stage model (MSE = 10.29, R² = 0), DNN (target data set (MSE = 10.77, R² = 0)), MAML (MSE = 43.45, R² = 0)

SRRC#13

Train

Successful models:MMN(MSE = 0.44, R² = 0.89) > Four-stage model (MSE = 0.60, R² = 0.86)

Failed models: MAML (MSE = 5.05, R² = 0), DNN (target data set (MSE = 14.36, R² = 0)), DNN (all data set (MSE = 41.61, R² = 0)), Fine-tune (MSE = 98.12, R² = 0)

SRRC#2

Test

Successful models:MMN (MSE = 0.49, R² = 0.92) > Four-stage model (MSE = 0.50, R² = 0.91) > MAML (MSE = 1.87, R² = 0.68)

Failed models: DNN (target data set (MSE = 21.94, R² = 0)), Fine-tune (MSE = 34.67, R² = 0), DNN (all data set (MSE = 35.37, R² = 0))

SRRC#4

Test

Successful models: MMN (MSE = 0.07, R² = 0.99) > Four-stage model (MSE = 0.62, R² = 0.92) > Fine-tune (MSE = 1.84, R² = 0.76) > DNN (all data set (MSE = 3.94, R² = 0.48)) > MAML (MSE = 5.87, R² = 0.22)

Failed models:, DNN (target data set (MSE = 15.59, R² = 0))