Jawaharlal Nehru Technological University Anantapur 2010-1st Sem B.Tech Code :R7411009 2 IV (R07) Regular s, ember/ember ARTIFICIAL NEURAL NETWORKS - Question Paper

Code :R7411009 2
IV B.Tech I semester (R07) Regular Examinations, November/December 2010
ARTIFICIAL NEURAL NETWORKS
(Electronics & Instrumentation Engineering)
Time: three hours Max Marks: 80
ans any 5 ques.
All ques. carry equal marks
1. (a) With neat diagram discuss biological neuron.
(b) discuss neuron modeling for arti¯cial neural sytems.
2. (a) discuss perceptron leaning rule with the help of necessary diagram.
(b) Implement the perceptron rule training for the above architecture using f(net)=sgn(net),
c=1, and the subsequent data specifying the initial weights w1, and the 2 training pairs
W1 = £ 0 one 0 ¤T
;
(X1 = £ two one ¡1 ¤T
; d1 = ¡1)
(X2 = £ 0 ¡1 ¡1 ¤T
; d2 = 1)
Repeat the training sequence (x1,d1) , (x2,d2) unit 2 accurate responses in a row are
achieved. List the net K values found during training.
3. (a) Draw and discuss the radical basis function networks.
(b) discuss the single layer perceptron architecture.
4. What is Madaline ?Draw the architecture of Madaline net? discuss the MRI algorithm used
in a Madaline net.
5. (a) discuss different applications of counter propagation.
(b) discuss self organizing feature maps.
6. (a) discuss the continuous Hop¯eld net with its architecture.
(b) explain in detail the energy function used in the discrete Hop¯eld net.
7. Draw the architecture of a BAM network and explain in detail the training algorithm.
8. (a) discuss image compression using direct solution method based neural network.
(b) discuss about solving optimization issues.

Earning:   Approval pending.