This is an example of perceptron learning and testing.  The idea is to have
Persim be able to tell which side of a line a particular point is.

The only difference between example2 (this directory) and example1 is the
random X & Y coordinates for the ten points (and the alpha used).

The boundary line used is:

	y = -1/2X + 1

Ten random points were generated (input1.asc - input10.asc).  An extra input
node of value 1 was added to each input set instead of having an output
node threshold.

Desired1.asc - desired10.asc is the expected answer for each corresponding
input set (input1.asc with expected output desired1.asc).

Weights.asc is the file containing the weights.  This file get updated during
training and used during testing.  This file is initially created by the
first training session with Persim.

Output1 - output10 is the output for each session with Persim.

Runit is a shell script to automate the training and testing of Persim.  It
consists of 2 FOR loops.  The first FOR loop creates batch files and runs
Persim to train and the second FOR loop creates batch files and runs Persim
for testing.

