B.TECH. DEGREE EXAMINATION, MAY 2016
Eighth Semester
Branch : Computer Science and Engineering/Information Technology
CS 010 805 G02/IT 010 805 G05 -NEURAL NETWORKS (Elective IV) (CS, IT)
(New Scheme 2010 Admission onwards)
[Regular/Supplementary]
Time Three Hours
Maximum : 100 Marks
Part A
Answer all questions.
Each question carries 3 marks.
- What is the difference between human intelligence and machine intelligence ?
- What are multilayer perceptrons ?
- What are the applications of Radial Basis Function networks ?
- What are the various ways to realize competition in neural networks ?
- What is pattern association ?
(5 x 3 = 15 marks)
Part B
Answer all questions.
Each question carries 5 marks.
- Why we need pattern classification ?
- Explain temporal stability.
- How to train RBF networks ? Explain.
- What is a counter propagation network ?
- Give an example for hetero associative networks.
(5 x 5 = 25 marks)
Part C
Answer all questions.
Each question carries 12 marks.
- Compare the performance of a computer and that of a biological neural network in terms of speed of processing, size and complexity, storage, fault tolerance and control mechanism.Or
- Explain McCulloch-Pitts model of neuron.
- Briefly explain Back propagation algorithm in detail.Or
- Derive the Back propagation algorithm for regression.
- What is under fitting and over fitting ? Explain prevention mechanisms for the same.Or
- Explain learning with momentum and conjugate gradient learning.
- Explain Maxnet fixed weight competitive networks.Or
- Explain the application and architecture of full counter propagation network.
- Briefly explain the algorithm for hopfield networks.Or
- What is the purpose of using bidirectional associative memory ? Explain.
(5 x 12 = 60 marks)