BSD 3204  BISF 3203  MACHINE LEARNING.

UNIVERSITY EXAMINATIONS: 2020/2021
EXAMINATION FOR BACHELOR OF SCIENCE IN INFORMATION
SECURITY AND FORENSICS
BSD 3204/BISF 3203: MACHINE LEARNING
FULL TIME/ PART TIME
ORDINARY EXAMINATION
DATE: DECEMBER, 2021 TIME: 2 HOURS
INSTRUCTIONS: Question ONE IS COMPULSORY, Choose TWO OTHER Questions

QUESTION ONE (20 MARKS) COMPULSORY
a). Explain the following main categories of machine learning, giving an example of
algorithm that falls under the category in each case
i). Supervised machine learning
ii). Unsupervised machine learning
iii). Reinforcement learning (6 Marks)
b). The goal of machine learning is to generate a model that is as a result of looking at the
data and learning the pattern present in the data, however a model faces two challenges,
underfitting and overfitting.
i). Explain the terms underfitting and overfitting
b). Describe how this challenges can be overcome during model development
(8 Marks)
c). Explain the function of learning rate in gradient descent (3 Marks)
d). Highlight three weakness of k-means clustering algorithm (3 Marks)
QUESTION TW0 (15 MARKS)
a). Discuss the main advantages of machine learning in the area of artificial intelligence
(5 Marks)
b). Identify and discuss five applications of machine learning in business (10 Marks)
QESTION THREE (15 MARKS)
a). Discuss the following concepts as applied in decision trees
i). Gini index
ii). Information gain ratio
iii). Entropy (3 Marks)
b). Data preprocessing is an a very important step in machine learning and data mining,
describe four activities that can be carried out during data preprocessing (8 Marks)
c). Using an example, describe how Bayesian machine learning algorithm can be applied in
real life problems. (4 Marks)
QUESTION FOUR (15 MARKS)
a). Describe how the winnow algorithm differs from the perceptron algorithm
(6 Marks)
b). Explain the functions of the following python libraries
i). Numpy
ii). Pandas
iii). Matplotlib
iv). Seaborn (4 Marks)
c). Explain the following terms as used in machine learning
i). Cross validation
ii). Dropout
iii). Regularization
iv). Standardization
v). Precision (5 Marks)

(Visited 133 times, 1 visits today)
Share this:

Written by