Cell Recognition Using BP Neural Network Edge Computing

Table 1

BPNN parameter settings.

Parameter	Value
Learning rate	0.3
Maximum training times	100
Accuracy required for training	0
Minimum gradient requirements	1.12E-10
Momentum factor	0.9
The activation function of the output layer	Purelin linear function
Number of network layers	4
Weight change increment	1.2
Reduction of weight change	0.5
Initial weight change	0.07
Loss function	Quadratic mean square function
Activation function	Sigmoid function