Scribd
Upload
159 views2 pages

CS2108 - 1 - Sem MAY 2018

This document appears to be an exam for a pattern recognition techniques course. It contains: 1) A list of 10 short answer questions covering topics like apriori and posterior probability, pattern classification, Bayes' rule, Parzen windows, linear separability, unsupervised learning approaches, regular languages, tree grammars, constructing a weight matrix for an associative memory, and sigmoid activation functions. 2) A list of 5 longer answer questions worth 16 marks each, covering topics such as comparing PR approaches, statistical vs structural PR, Bayesian parameter estimation, K-nearest neighbors estimation, hierarchical clustering, C-means clustering, CYK parsing, and neural network architectures. 3) The exam is worth a total

Uploaded by

Jeba Nithi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
159 views2 pages

CS2108 - 1 - Sem MAY 2018

This document appears to be an exam for a pattern recognition techniques course. It contains: 1) A list of 10 short answer questions covering topics like apriori and posterior probability, pattern classification, Bayes' rule, Parzen windows, linear separability, unsupervised learning approaches, regular languages, tree grammars, constructing a weight matrix for an associative memory, and sigmoid activation functions. 2) A list of 5 longer answer questions worth 16 marks each, covering topics such as comparing PR approaches, statistical vs structural PR, Bayesian parameter estimation, K-nearest neighbors estimation, hierarchical clustering, C-means clustering, CYK parsing, and neural network architectures. 3) The exam is worth a total

Uploaded by

Jeba Nithi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Reg. No.

Ph. D. (Course Work) EXAMINATION, MAY 2018

cs2108 - PATTERN RECOGNTTTON TECIINIQUES


Time: Three Hours Max. Marks: 100
Answer ALL Questions
PART-A(10x2:20Marks)
1. What is meant by apriori probability and posterior
probability?
2. Define: Pattern and Pattern class.
3. What does Baye's rule convey in PR domain?
4. What are Parzen windows?
5. Define: Linear separability
6. What do you mean by non-parametric approach of
unsupervised learning?
7. Give the conditions for a language to be regular.
8. Define: Tree grammar, Give example.
9. Given the training pairs. Construct the weight matrix 'W'
for Hetero associate memory.
E:(4 I o) &:(1 o o)
s2:(4 1 o) fi2 = (o r o)
sr=(2 o l) n3=(o o 1)
10. Why is sigmoidal activation function famous in ANN?

PART-B(5x16=80Marks)
11.a. Compare the various PR approaches and explain how they
are different in their functionality.

(oR)
Page I of3 llMACS2l0S
b. Differentiate statPR and syntPR in the context of b. Discuss the following:
model-based PR. (i) Content addressable memory
(ii) ARr
12.a. Discuss the Bayesian parameter estimation approach to aid
,f**r(
in classification.

(oR)
b. Explain the K-NN nonparametric estimation and highlight
on estimating P(wila) directly.

13.a. Describe the Hierarchical clustering procedure and analyse


the result of making the number of samples same in all
clusters.

(oR)
b. Give an algorithm for C-means clustering and how does
the Gaussian function changes affect the desired output?

I4.a. Given the grammar below, construct the CYK parsing


table and check for the string x : aabb.
S-+ABIBB;
A-+CClABla;
B-+BBlCAlb;
C-+BAlAAlb:
(oR)
b. Describe how structural unification is done using
Attributed graphs with example.

15.a. With a neat diagram explain the various architectures and


learning algorithms in Artificial neural networks.

(oR)

Page 2 of3 rlMACS2l0S Page 3 of 3 llMACS2l08

You might also like