Improved Machine Learning Method for Intracranial Tumor Detection with Accelerated Particle Swarm Optimization

Table 1

Parameters definition for ANN.

Parameters	Initial values
Activation functions	“Purelin” and “tansig”
Nodes in hidden layer	16
Learning model	Levenberg–Marquadrt backpropagation
Learning rate	0.005
Number of epochs	500
Mean square error	10^–5
Minimum error	0.002
Momentum rate	0.6
Training time	50 seconds