TYMCA (Part (Under Faculty of Engg.) Examination, 2017
Elective II FUZZY LOGIC AND ARTIFICIAL NEURAL NETWORK
Day and Date Tuesday, 16-5-2017 Total Marks 100
Time 10.30 a.m. to 1.30 p.m.
Instructions Figures to the right indicate full marks.
Q. 3.A) and Q. 5A) are compulsory.
MCQ/Objective Type Questions.
1. Choose the correct alternative. 20
A fuzzy set A contains an object x to degree this is a Degree
and the map a X → {Membership Degree} is called
A set function A membership function
Both a and b None of the above
The theory of fuzzy logic is based upon the notion of
graded membership and so are the function of cognitive.
Relative Absolute
Part of None of the above
The intersection between two crisp sets represents all those elements in the
universe that
Belongs to both sets Belongs to any on set
Both a and b None of the above
Fuzzy sets viewed as of the basic concepts of crisp sets.
Extension Generalization
Both a and b None of the above
The properties hold good for fuzzy relations.
Commutative Associativity
Idempotency None of the above
A fuzzy set works as a concept that makes it possible to treat fuzziness in a
manner.
Qualitative Quantitative
Both a and b None of the above
A fuzzy set wherein no membership function has its value equal to
is called subnormal fuzzy set.
One Two
Three None of the above
method employs the algebraic sum of the
individual fuzzy subsets of their union.
Center of sums Weighted average
Mean-max None of the above
is the process of conversion of a fuzzy quantity into a precise
quantity.
Fuzzification Defuzzification
Both a and b None of the above
10) Applications of the FLC are
Traffic control Missile control
Adaptive control All of the above
11) Sigmoidal function is function that varies gradually between
values 0 and 1 or 1 and 1.
Continuous Discrete
Logical None of the above
12) ANN structure can be represented by
Graph Directed Graph
Tree None of the above
13) Neural networks have shown remarkable progress in the recognition of visual
images
Handwritten characters Printed characters
Speech recognition All of the above
14) Activity of neurons in the hidden layer is determined by the activities of the
neurons in the
Input layer Output layer
Both a and b None of the above
15) A selection of tuning parameters are required for efficient
learning and design of stable BPN Network.
Momentum factor Sigmoidal function
Threshold value All of the above
16) A network is said to be network if no neuron in the output
layer is an input to a node in the same layer or in the preceding layer.
Feed forward Feedback
Lateral feedback None of the above
17) A neuron generates an output if the weighted sum of the input
the threshold value.
Exceeds Equal to
Less than None of the above
18) An associative memory belongs to the class of feedforward
neural network architecture.
Single layer Multi layer
Both a and b None of the above
19) A Hebb rule is widely used for finding the weights of neural
network.
Feedforward BPN
Associative memory None of the above
20) The BAM network performs associative searches for stored
stimulus responses.
Forward Backward
Both a and b None of the above
SECTION I
2. Write short note on any four.
Properties of fuzzy sets.
Operations on fuzzy relations.
Features of membership functions.
Defuzzification methods.
Multiattribute decision making.
3. Explain with block diagram architecture and operation of FLC system. 10
Explain in brief deffuzzification methods. 10
OR
Describe Fuzzy Rule based system. 10
SECTION II
4. Write short note on any four.
Learning methods.
Medaline networks.
MeCulloch-Pitts Neuron.
Learning difficulties and improvements.
Associative memory.
5. Explain working of multilayer feed forward back propagation network. 10
Explain in brief selection of various parameters in BNP. 10
OR
Explain concept of Auto associative memory network. 10