Neural Network Toolbox

Simulation (sim)

To show how sim works we examine a simple problem.

Suppose we take a perceptron with a single two-element input vector, like that discussed in the decision boundary figure. We define the network with

• net = newp([-2 2;-2 +2],1);
  

As noted above, this gives us zero weights and biases, so if we want a particular set other than zeros, we have to create them. We can set the two weights and the one bias to -1, 1 and 1 as they were in the decision boundary figure with the following two lines of code.

• net.IW{1,1}= [-1 1];
  net.b{1} = [1];
  

To make sure that these parameters were set correctly, we check them with

• net.IW{1,1}
  ans =
      -1     1
  net.b{1}
  ans =
       1
  

Now let us see if the network responds to two signals, one on each side of the perceptron boundary.

• p1 = [1;1];
  a1 = sim(net,p1)
  a1 =
       1
  

and for

• p2 = [1;-1]
  a2 = sim(net,p2)
  a2 =
       0
  

Sure enough, the perceptron classified the two inputs correctly.

Note that we could present the two inputs in a sequence and get the outputs in a sequence as well.

• p3 = {[1;1] [1;-1]};
  a3 = sim(net,p3)
  a3 = 
      [1]    [0]
  

You may want to read more about sim in Advanced Topics.

Creating a Perceptron (newp)

Initialization (init)

© 1994-2005 The MathWorks, Inc.