Ann 2023-2023

MAIN UNIVERSITY EXAMINATIONS 2022/2023
EXAMINATION FOR THE DEGREE OF BACHELOR OF SCIENCE IN SOFTWARE
ENGINEERING
SPC 2409 Artificial Neural Networks Year IV Semester I

Date: Monday, 5th December 2022 Time: 11.00am – 1.00pm
INSTRUCTIONS
i. This EXAMINATION has TWO sections.

ii. SECTION A is compulsory. Answer ALL questions in this section.
iii. Answer ANY TWO questions in SECTION B.
SECTION A
QUESTION ONE (30 MARKS)
a) Distinguish between each of the following terms as used in Neural Networks

(i) Supervised and unsupervised learning (2 marks)
(ii) Pattern classification and pattern association (2 marks)
(iii) Feed forward network and recurrent network (2 marks)
(iv) Artificial neural networks and Biological neural networks (2 marks)
b) The basis operation of an artificial neuron involves summing its weighted input
signal and applying an activation function to determine whether it will ‘fire’ or not
(i) Explain what you understand by the term ‘activation function’ (2 marks)
(ii) Using appropriate illustrations, explain each of the following activation
functions
i. Binary Step Function (3 marks)
ii. Bipolar Sigmoid Function (3 marks)
c) Artificial neural networks (ANNs) are modelled from biological neurons. Outline
the biological properties that are borrowed into ANNs (6 marks)
d) Briefly explain any application areas of Neural networks (4 marks)
Page 1 of 3
e) You have been hired as the data scientist in the meteorological department. Your
first assignment involves the use of neural networks for time series prediction
specifically weather forecasting. Explain the techniques that you would apply for
this task and justify why (4 marks)
QUESTION TWO (20 MARKS)
a) Distance metrics play a significant role in neural network problems in determining
the similarity and diversity of sample sets. Using appropriate examples, explain
each of the following distance methods
i. Euclidean distance (2 marks)
ii. Jaccard index (2 marks)
iii. Hamming distance (2 marks)
b) Self-organizing maps are widely used in clustering problems of neural networks.
Explain how the Kohonen self-organizing map works (4 marks)
c) You have been hired in a Big Data company involved in prediction of traffic
patterns. Explain the techniques that you would apply for this task and justify why
(5 marks)
d) Compare and contrast an artificial neural network and the biological neuron
(5 marks)
QUESTION THREE (20 MARKS)
a) Using an illustration, explain the key components of a simple neural network
(4 marks)
b) A network topology is the arrangement of a network along with its nodes and
connecting line. Explain two main categories of network topology (4 marks)
c) ART network is a vector classifiers which accepts an input vector and classifies it,
with respect to vector classification explain three operations of ART (6 marks)
d) Differentiate between auto associative memory and hetero associative memory
(4 marks)
e) Boltzmann machine is an asymmetric coupled random feedback binary unit
neural network, which includes a visible layer and multiple hidden
layers. Explain two importance of Boltzmann Machine (2 marks)
Page 2 of 3
QUESTION FOUR (20 MARKS)
a) Artificial neural networks differs with biological neural networks because of the
few characteristics. Explain three characteristic that defines neural networks
(6 marks)
b) With a clear structure interpret the following function as used in artificial neural
networks
Yin=x1.w1+x2.w2+x3.w3…xn.wn (5 marks)
c) Biological neural networks is a series of interconnected neurons whose activation
defines a recognizable linear pathway. Explain the function of the following
components
a. Dendrites (2 marks)
b. Axon (2 marks)
c. Soma (2 marks)
d) Artificial intelligence is the ability of a computer or a robot controlled by a
computer to do tasks that are usually done by humans because they require
human intelligence and discernment. Explain three branches of artificial
Intelligence (3 marks)
QUESTION FIVE (20 MARKS)
a) What is Multi-layer feed forward networks? What is the importance of hidden
and output layers in it? (7 marks)
b) A recurrent neural network (RNN) is a type of artificial neural network which
uses sequential data or time series data. With an aid of a diagram explain fully
recurrent network and Jordan network (6 marks)
c) Suppose that a credit card company decided to deploy a new system for
assessing credit worthiness of its customers. The new system is using a feed-
forward neural network with a supervised learning algorithm. Suggest what the
bank should have before the system can be used? Discuss problems associated
with this requirement (5 marks)
d) Explain the following terms (2 marks)
a. Neurons
b. Perception
Page 3 of 3

Ann 2023-2023

Uploaded by

Copyright:

Available Formats

Ann 2023-2023

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ann 2023-2023

Uploaded by

Copyright:

Available Formats

MAIN UNIVERSITY EXAMINATIONS 2022/2023

EXAMINATION FOR THE DEGREE OF BACHELOR OF SCIENCE IN SOFTWARE

SPC 2409 Artificial Neural Networks Year IV Semester I

i. This EXAMINATION has TWO sections.

a) Distinguish between each of the following terms as used in Neural Networks

You might also like