Table 11 Statistic metrics obtained by W-GRNN model to forecast the EC parameter for all combinations.

From: An improved adaptive neuro fuzzy inference system model using conjoined metaheuristic algorithms for electrical conductivity prediction

Model

Criteria

Combination

Combo 1

Combo 2

Combo 3

Combo 4

Wavelet-Demy Test

R

0.811

0.785

0.797

0.790

RMSE

214.160

224.622

220.293

225.909

MAE

166.059

181.860

179.529

185.436

RAE

0.598

0.655

0.647

0.668

MAPE

9.055

9.727

9.442

9.677

E

0.634

0.597

0.612

0.592

IA

0.879

0.869

0.881

0.881

PI

0.937

0.993

0.974

0.996

Wavelet-Bior Test

R

0.810

0.791

0.792

0.789

RMSE

219.122

221.946

222.328

225.607

MAE

171.231

180.733

180.477

184.291

RAE

0.617

0.651

0.650

0.664

MAPE

9.238

9.560

9.598

9.676

E

0.617

0.607

0.605

0.594

IA

0.876

0.877

0.875

0.878

PI

0.960

0.987

0.988

1.000

Wavelet-Demy Train

R

0.933

0.959

0.974

0.986

RMSE

251.635

200.526

159.135

115.105

MAE

185.355

148.077

114.561

77.823

RAE

0.342

0.273

0.211

0.144

MAPE

10.074

7.874

5.902

3.837

E

0.857

0.909

0.943

0.970

IA

0.956

0.973

0.984

0.992

PI

1.000

0.868

0.756

0.639

Wavelet-Bior Train

R

0.957

0.969

0.966

0.982

RMSE

204.698

176.366

182.512

134.219

MAE

146.654

128.769

133.566

93.273

RAE

0.271

0.238

0.246

0.172

MAPE

7.787

6.737

7.003

4.669

E

0.906

0.930

0.925

0.959

IA

0.972

0.980

0.978

0.989

PI

1.000

0.920

0.939

0.775

  1. Significant values are in bold.
Back to article page