Professional Documents
Culture Documents
Module 2 Hebb Net
Module 2 Hebb Net
Module 2 Hebb Net
wi(new) = wi(old) + ∆w
∆w weight change
ARCHITECTURE
HEBBIAN LEARNING-TRAINING
ALGORITHM
Step 0: First initialize the weights. Basically in this
network they may be zero, i.e., w= 0 for i= 1to n
where "n" may be the total number of input neurons.
Step 1: Steps 2-4 have to be performed for each input
training vector and target output pair, s: t.
Step 2: Input units activations are set.
x1 x2 y(target)
1 1 1
1 -1 -1
-1 1 -1
-1 -1 -1
Initialize :
w1=w2=0 and bias, b=0
∆b=y
SOLVED PROBLEM-1
inputs target Change in Change in bias New New New
weights weight weight bias
1 1 1 1 1 1 1 1 1
1 -1 -1 -1 1 -1 0 2 0
-1 1 -1 1 -1 -1 1 1 -1
-1 -1 -1 1 1 -1 2 2 -2
x1 x2 y (target)
1 1 -1
1 -1 1
-1 1 1
-1 -1 -1
SOLVED PROBLEM-2
1 1 -1 -1 -1 -1 -1 -1 -1
1 -1 1 1 -1 1 0 -2 0
-1 1 1 -1 1 1 -1 -1 1
-1 -1 -1 1 1 -1 0 0 0
EXERCISE PROBLEMS
1. Design a Hebb net to implement OR function (consider
bipolar inputs and targets).
EXERCISE PROBLEMS
1. Using the hebb rule, find the weights required to
perform the following classifications:
vectors (1 1 1 1) and (-1 1 -1 -1) are the members of class
(with target value 1) ;
vectors (1 1 1 -1) and (1 -1 -1 1) are the members of class
(with target value -1)
x1 x2 x3 x4 t
1 1 1 1 1
-1 1 -1 -1 1
1 1 1 -1 -1
1 -1 -1 1 -1
APPLICATION BASED QUESTION
Classify the two-dimensional input pattern using Hebb
neural network (letters I-J pattern)