Machine Learning - R_23 - JNTUK, Model Paper - 2

JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY KAKINADA

B.Tech III Year II Semester – R23

Course: MACHINE LEARNING

Time: 3 Hours Max Marks: 70

---

SECTION – I (50 Marks)

Answer ALL questions
(Each question carries 10 marks: a) 5 + b) 5)

UNIT – I : Introduction to Machine Learning

1.

a) Define Machine Learning paradigms. Explain supervised and unsupervised learning with real-world examples. (5M)
b) Explain Types of Data used in Machine Learning. How does data type affect algorithm selection? (5M)

2.

a) Explain Data Representation in Machine Learning. Discuss different forms of representing data. (5M)
b) Describe Model Selection and Model Learning. Explain their role in building an effective ML system. (5M)

UNIT – II : Nearest Neighbor–Based Models

3.

a) Explain Proximity Measures and Similarity Measures used in nearest neighbor models. (5M)
b) Apply Manhattan and Euclidean distance measures to find the nearest neighbor for a given test instance. (5M)

4.

a) Explain the K-Nearest Neighbor (KNN) Classifier algorithm with a flow diagram. (5M)
b) Discuss KNN Regression and explain how it differs from KNN classification. (5M)

UNIT – III : Decision Trees & Bayesian Models

5.

a) Explain the properties of decision trees and their advantages and disadvantages. (5M)
b) Explain Regression using Decision Trees with an example. (5M)

6.

a) Explain Bayes’ Theorem and its importance in classification problems. (5M)
b) Explain Multi-class classification using Naive Bayes classifier with a suitable example. (5M)

UNIT – IV : Linear Discriminants & Neural Networks

7.

a) Explain Perceptron classifier and its limitations. (5M)
b) Describe Logistic Regression and explain how it differs from linear regression. (5M)

8.

a) Explain Support Vector Machines (SVM) for linearly separable cases with a diagram. (5M)
b) Explain Backpropagation algorithm used for training Multi-Layer Perceptrons. (5M)

UNIT – V : Clustering

9.

a) Explain Partitioning of Data in clustering. Discuss challenges in clustering large datasets. (5M)
b) Apply K-Means algorithm to cluster the following data into two clusters. (5M)

10.

a) Explain Agglomerative Clustering with suitable examples. (5M)
b) Describe Spectral Clustering and mention its advantages over traditional clustering methods. (5M)

SECTION – II (20 Marks)

(Each question carries 2 marks)

1. Define Machine Learning.
2. What is reinforcement learning?
3. Define feature engineering.
4. What is Hamming distance?
5. Define overfitting.
6. What is entropy?
7. What is kernel trick?
8. Define activation function.
9. What is fuzzy clustering?
10. Define expectation maximization.