Table 9 Performance comparison with different optimizers.

From: Multi-kernel inception-enhanced vision transformer for plant leaf disease recognition

Optimizer

Training Acc

Training loss

Validation Acc

Validation Loss

Apple dataset41

 SGD

0.4638

0.9856

0.3891

1.0374

 RMSProp

0.9982

0.1178

0.9968

0.1251

 Adam

0.9972

0.3728

0.9923

0.3732

 Adamax

0.8147

0.8556

0.7023

0.8703

 Adadelta

0.8238

0.5591

0.7176

0.8390

 Nadam

0.8268

0.5621

0.7248

0.8232

 Ftrl

0.8562

0.5394

0.7451

0.8132

Rice dataset34

 SGD

0.8692

0.5346

0.7318

0.6146

 RMSProp

1.000

0.3393

0.9927

0.3567

 Adam

1.000

0.3328

0.9970

0.3528

 Adamax

0.9168

0.4285

0.8527

0.5128

 Adadelta

0.8571

0.5348

0.8038

0.5793

 Nadam

0.8179

0.5249

0.7748

0.6173

 Ftrl

0.8027

0.5294

0.7819

0.6123

Ibean dataset36

 SGD

0.4152

1.0348

0.3750

1.0877

 RMSProp

0.9826

0.3290

0.9294

0.4241

 Adam

0.9965

0.3028

0.9702

0.4028

 Adamax

0.4152

1.0348

0.3750

1.0877

 Adadelta

0.4152

1.0348

0.3750

1.0877

 Nadam

0.4152

1.0348

0.3750

1.0877

 Ftrl

0.4152

1.0348

0.3750

1.0877

Cassava dataset42

 SGD

0.7682

0.6251

0.5025

0.7396

 RMSProp

0.9046

0.4376

0.7381

0.6451

 Adam

0.9247

0.4228

0.7651

0.6328

 Adamax

0.7187

0.6429

0.4627

0.7914

 Adadelta

0.7097

0.6452

0.4607

0.8014

 Nadam

0.7285

0.6104

0.4821

0.7552

 Ftrl

0.7047

0.6625

0.4663

0.7936

PlantVillage dataset14

 SGD

0.8732

0.4753

0.8271

0.5129

 RMSProp

0.9952

0.1393

0.9918

0.2178

 Adam

0.9973

0.1347

0.9942

0.1836

 Adamax

0.9263

0.4621

0.8575

0.4726

 Adadelta

0.8357

0.4961

0.7817

0.6129

 Nadam

0.9136

0.3961

0.8537

0.4327

 Ftrl

0.8436

0.4853

0.7971

0.5326

Back to article page