00000EC467121903

Pages: 2

Reg No.: Name:

APJ ABDUL KALAM TECHNOLOGICAL UNIVERSITY

Seventh Semester B.Tech Degree Examination (Regular and Supplementary), December 2020

Max. Marks: 100

1 a)
b)
2 a)
b)
3 a)
b)
4 a)
b)
5 a)
b)
6 a)
b)
7 a)
b)

Course Code: EC467
Course Name: PATTERN RECOGNITION

PARTA

Answer any two full questions, each carries 15 marks.
Explain the various applications of pattern recognition systems.
Obtain the discriminant function for Bayes classifier if the feature vector
distribution is Gaussian with different means and a fixed diagonal covariance
matrix.
Explain the Bayesian parameter estimation technique.
Describe the significance of Gaussian mixture models in classifier design.
For a two category Bayes classifier, the loss function is given by 41;=0.1, 221 =
1, 2121, 222 = 0.2. The categories are equally likely. Obtain the decision rule.
Explain Fisher discriminant analysis for dimensionality reduction.

PART 1
Answer any two full questions, each carries 15 marks.

Explain K Nearest Neighbour method for density estimation.

Explain the perceptron model for classification.

Explain support vector machines and how it achieves maximum margin
classification.

Define overfitting and its drawback.

Define the various impurity measures used in test selection while constructing a
decision tree.

Explain gradient descent algorithm and state perceptron convergence theorem.

PART C
Answer any two full questions, each carries 20 marks.
What is bagging approach in ensemble classifier?

Explain the classification capabilities of a two layer perceptron with necessary

illustrations.

Page lof 2

Duration: 3 Hours

Marks
(5)
(10)

(8)
(7)
(5)

(10)

(10)
(5)
(10)

(5)
(8)

(7)

(7)
(8)