Table 7 Sensitivity analysis of the GRU network with different numbers of hidden neurons.

From: Performance investigation of Xanthan gum polymer flooding for enhanced oil recovery using machine learning models

Number of hidden neurons

Data type

\(\:{\varvec{R}}^{2}\)

MSE

RMSE

5

Training

Validation

All Data

0.973576

0.97779

0.974422

7.101665

6.692484

7.019479

2.659229

2.583303

2.6494

7

Training

Validation

All Data

0.981347

0.990384

0.983162

5.56097

2.92758

5.032041

2.259143

1.703374

2.243221

9

Training

Validation

All Data

0.983761

0.991326

0.985280

3.817912

2.67782

3.588919

1.915558

1.617833

1.894444

11

Training

Validation

All Data

0.983469

0.993254

0.985434

4.352224

2.056572

3.891131

2.066194

1.42692

1.972594

14

Training

Validation

All Data

0.98206

0.989182

0.9834904

5.160986

3.19948

4.767008

2.230541

1.788295

2.183347

  1. Values in bold indicate the optimal model performance.
Back to article page