G192120 Pages:2
Reg No.: Name:
APJ ABDUL KALAM TECHNOLOGICAL UNIVERSITY
SEVENTH SEMESTER B.TECH DEGREE EXAMINATION(R&S), DECEMBER 2019
Course Code: EC467
Course Name: PATTERN RECOGNITION
Max. Marks: 100 Duration: 3 Hours
PARTA
Answer any two full questions, each carries 15 marks. Marks
1 a) Differentiate between supervised and unsupervised learning with examples. (6)
b) Fora Bayesian classifier, obtain the discriminant function and decision surface (9)
for multivariate Gaussian distributions having same covariance matrix and
distinct means.
2 a) Explain expectation maximization algorithm. (7)

b) Draw the block diagram of a pattern recognition system. Explain the terms i) (8)
features ii) training and 111) testing.
3 9) Describe principal component analysis (PCA) as a technique for dimension (10)
reduction of features.

b) How mixture models are created using Gaussian densities? (5)

PART B
Answer any two full questions, each carries 15 marks.
4 a) Discuss the non-parametric Parzen window method of estimating an unknown (9)
probability density function.
b) 1150 different types of activation functions used in perceptron models. (6)
5 2) Explain k-Nearest-Neighbour method for estimating an unknown probability (8)
density function.
b) Explain the terms 1) splitting of nodes 11) attribute selection 111) over-fitting, and (7)
iv) pruning in the context of decision trees.
6 9) Illustrate the Perceptron algorithm for two linearly separable classes. (8)

b) Describe a Support Vector Machine. (7)

Page lof 2