Skip to main content

Table 2 Values of the hyperparameters for the neural networks

From: Deep learning for detecting herbicide weed control spectrum in turfgrass

Deep learning architecture

Optimizer

Base learning rate

Learning rate policy

Batch size

Training epochs

GoogLeNet

Adam

0.0003

StepLR

48

60

MobileNet-v3

Adam

0.0001

StepLR

48

60

ShuffleNet-v2

SGD

0.001

LambdaLR

48

60

VGGNet

Adam

0.0001

StepLR

48

60

  1. SGD stochastic gradient descent