Neural Network Toolbox

Network Initialization

The network has only one neuron, as only one output value of the signal T is being generated at each time step. This neuron has five inputs, the five delayed values of the signal T.

The function newlin creates the network shown above. We use a learning rate of 0.1 for incremental training.

• lr = 0.1;
  delays = [1 2 3 4 5];
  net = newlin(minmax(cat(2,P{:})),1,delays,lr);
  [w,b] = initlin(P,t)
  

Applin2: Adaptive Prediction

Network Training