Pattern Recognition

1.Divide data for validating:
validating data： that the network is generalizing and to stop training before overfitting.
2.About number of neurons and layers:
More neurons require more computation, and they have a tendency to overfit the data when the number is set
too high, but they allow the network to solve more complicated problems. More
layers require more computation, but their use might result in the network solving
complex problems more efficiently. To use more than one hidden layer, enter the
hidden layer sizes as elements of an array in the patternnet command.
3.The pattern recognition network uses the default Scaled
Conjugate Gradient (trainscg) algorithm for training.