Skip to main content
Journal of Big Data

Table 4 Determination of optimal machine learning models before and after heterogeneity

From: Detecting heterogeneity parameters and hybrid models for precision farming

ML models

High ranking variables

Metric validations before heterogeneity

Metric validations after heterogeneity

Absolute change

Percentage change

MAPE

MSE

R2

MAPE

MSE

R2

Ridge (control)

15

14.64603

83.92337

0.6932309

13.45105

82.84078

0.6971881

1.194980

8.159071

25

11.50656

56.4660

0.7935971

12.63606

75.31612

0.7246934

1.129500

− 9.81614

35

10.0306

48.23541

0.8236828

12.00500

70.29434

0.7430497

1.974400

− 19.68380

45

9.657189

44.48745

0.8373829

11.95927

69.44397

0.7461581

2.302081

− 23.83800

Random forest

15

2.458969

9.910512

0.9637737

9.885843

67.35215

0.7538052

7.426874

− 302.03200

25

2.337353

9.010273

0.9670644

7.909333

47.21578

0.8274099

5.571980

− 238.38800

35

2.174667

7.790909

0.9715216

7.663343

45.15805

0.8349317

5.488676

− 252.39200

45

2.125891

7.330011

0.9732063

7.588079

44.39000

0.8377405

5.462188

− 256.93600

Support vector machine

15

8.614626

45.25618

0.8347612

11.77207

77.48160

0.7169731

3.157444

− 36.65210

25

7.980399

35.80985

0.8691446

11.15354

71.12697

0.7401082

3.173141

− 39.76170

35

7.568951

34.00095

0.8757802

10.89938

68.85807

0.7484105

3.330429

− 44.00120

45

7.351331

32.38644

0.8816661

10.62685

66.33326

0.7575719

3.275519

− 44.55680

Bagging

15

12.25897

74.29053

0.7284423

11.30002

66.52011

0.7568458

0.958950

7.822440

25

9.778194

47.33173

0.8269861

10.62821

57.44370

0.7900233

0.850016

− 8.69298

35

8.413645

36.41955

0.8668739

9.417039

48.41542

0.8230248

1.003394

− 11.92580

45

8.151903

33.65611

0.8769752

8.983211

45.01187

0.835466

0.831308

− 10.19770

Boosting

15

8.168942

142.4542

0.5310293

13.09470

217.8164

0.3416015

4.925758

− 60.29860

25

8.697362

136.3236

0.5543729

13.16813

215.4273

0.346658

4.470768

− 51.40370

35

8.183671

140.1463

0.5368431

12.78951

208.6947

0.3629861

4.605839

− 56.28080

45

8.203304

134.0864

0.5569358

8.228835

135.3237

0.5510545

0.025531

− 0.311230

LASSO

15

14.39656

101.8853

0.6275736

12.27376

74.04000

0.7293580

2.122800

14.74519

25

10.82264

52.90467

0.806615

11.65852

67.91064

0.751763

0.835880

− 7.72344

35

8.977735

37.69348

0.8622172

11.60206

67.40559

0.7536091

2.624325

− 29.23150

45

8.149872

31.57626

0.8845778

11.52189

66.86088

0.7556002

3.372018

− 41.37510

Elastic Net

15

13.12778

78.47416

0.7131497

12.31004

73.09066

0.7328282

0.817740

6.22908

25

9.485387

41.90456

0.8468243

11.72084

68.13366

0.7509478

2.235453

− 23.56730

35

9.051548

37.81546

0.8617713

11.64224

67.48376

0.7533234

2.590692

− 28.62150

45

8.191381

32.53884

0.8810592

11.66734

67.30154

0.7539895

3.475959

− 42.43430

Back to article page