UPTU B.Tech Pattern Recognition ECS 074 Sem 7_2011-12.pdf ...

0 downloads 347 Views 816KB Size Report
There was a problem previewing this document. Retrying... Download. Connect more apps... Try one of the apps below to op
Printed Pages-3

ECSO74

(Following Paper ID and Roll No. to be filled in your Answer Book

B.Tech.

(sEM. VX) THEORY EXAMTNATTON 201 1_r2 PATTERN RECOGNITION Time

: 3 Hours

Note

:-

1.

Tbtal Marlcs

(i)

Attempt all questions.

(iD

Make suitable assumption ifrequired.

Attempt any two parts

(a)

:

:

100

(10x2:20)

What do you mean by pattern recognition ? Explain. Describe design principles of pattern recognition system

with an example.

(b) (D

What do you mean by learning and adaptation

?

L,xplaln.

(ii) (c)

Write short note on pattern recognition approaches.

Explain the following and discuss their significance in pattern recognition with suitable example

(D (i,

:

Mean and Covariance

Chi Square Test.

ECSoT4|KIH-26642

[Turn Over

2.

Attempt any two parts

:

(10x2:20)

(a) What is Bay's Theorem ? Explain. Also discuss Bay's Classifier using some example in detail.

(b)

What is a discriminant function ? Discuss it in detail. In a

two class problem, the likelihood ratio is given

as

follows

:

p(xlC,) / p(xlC,) Write the discriminant function in terms of the likelihood ratio.

(c)

Describe the following with suitable example

:

(D NormalDensityFunction (iD UtilityTheory. 3.

Attempt any two parts

:

(10x2=20)

(a) What do you mean by dimension

reduction ? Discuss

Principal Component Analysis (PCA) algorithm for dimension reduction.

(b)

Write short notes on the following

(D (il) (c) 4.

:

Maximum-Likelihoodestimation. Expectation-maximization

Write a short note on Hidden Markov Model (HMM).

Attempt any two parts

:

(a) What do you meanby fuzzy decision

(10x2:20) making

?

Also discuss

the fuzzy classifi cation using suitable example.

(b) Write an algorithm for K-Nearest Explain.

ECS0T4tKlH-26642

neighbor estimation.

(c)

Write

(i)

a short note on the

following

:

Parametric vs. non-parpmetric pattern recognition methods.

(ir)

5. .

Parzen windows.

Write short notes on two ofthe following

(a)

\

:

(10x2=20)

What do you mean by supervised leaming and unsupervised

learning ? Explain. Discuss any unsupervised learning algorithm with some example.

(b)

What do you mean by clustering ? Explain. Discuss Kmeans clustering algorithm with suitable example.

(c)

Write short notes on the following

:

(D Clusteringvs.classification (iD Clustervalidation.

ECS0T4tKtH-26642

24625