YEAR OF

CATEGORY L T P CREDIT
INTRODUCTION
CST444 SOFT COMPUTING
PEC 2 1 0 3 2019

Preamble: This course enables the learners to understand the concepts of Soft Computing
techniques and its applications. It covers Artificial Neural Networks, operations and models of fuzzy
logic, genetic algorithms and multi objective optimization techniques. This course helps the students
to develop algorithms and solutions for different real world applications.
Prerequisite: NIL.

Mapping of course outcomes with program outcomes

CO1 Describe soft computing techniques and the basic models of Artificial Neural Network
(Cognitive Knowledge Level: Understand)

CO2 Solve practical problems using neural networks (Cognitive Knowledge Level: Apply)

CO3 Illustrate the operations, model and applications of fuzzy logic (Cognitive Knowledge
Level: Apply)

CO4 Illustrate the concepts of Genetic Algorithm (Cognitive Knowledge Level: Apply)

CO5 Describe the concepts of multi-objective optimization models and the need for using
hybrid soft computing approaches(Cognitive Knowledge Level: Understand)

Mapping of course outcomes with program outcomes

PO1 PO2 PO3 PO4 PO5 PO6 PO7 PO8 PO9 PO10 PO11 PO12

CO1

CO2

CO3

CO4

CO5
 Abstract POs defined by National Board of Accreditation

PO# Broad PO PO# Broad PO

PO1 Engineering Knowledge PO7 Environment and Sustainability

PO2 Problem Analysis PO8 Ethics

PO3 Design/Development of solutions PO9 Individual and team work

Conduct investigations of
PO4 complex problems PO10 Communication

PO5 Modern tool usage PO11 Project Management and Finance

PO6 The Engineer and Society PO12 Life long learning

Assessment Pattern

Continuous Assessment Tests

Bloom’s End Semester
Category Examination
Test 1 (%) Test 2 (%) Marks (%)

Remember 30 30 30

Understand 30 30 30

Apply 40 40 40

Analyze

Evaluate

Create

Mark Distribution

Total Marks CIE Marks ESE Marks ESE Duration

150 50 100 3
Continuous Internal Evaluation Pattern:
Attendance 10 marks
Continuous Assessment Tests(Average of Internal Tests 1 & 2) 25 marks

Continuous AssessmentAssignment 15 marks

Internal Examination Pattern

Each of the two internal examinations has to be conducted out of 50 marks. First series test shall be
preferably conducted after completing the first half of the syllabus and the second series test shall
be preferably conducted after completing remaining part of the syllabus. There will be two parts:
Part A and Part B. Part A contains 5 questions (preferably, 2 questions each from the completed
modules and 1 question from the partly completed module), having 3 marks for each question
adding up to 15 marks for part A. Students should answer all questions from Part A. Part B contains
7 questions (preferably, 3 questions each from the completed modules and 1 question from the
partly completed module), each with 7 marks. Out of the 7 questions, a student should answer any
5.

End Semester Examination Pattern:

There will be two parts; Part A and Part B. Part A contains 10 questions with 2 questions from each
module, having 3 marks for each question. Students should answer all questions. Part B contains 2
full questions from each module of which student should answer any one. Each question can have
maximum 2 sub-divisions and carries 14 marks.

Course Level Assessment Questions

Course Outcome 1 (CO1):
1. Describe the necessity of Activation function? Examine the various aspects of sigmoidal
activation function. List the drawbacks. Calculate the net output of the following neural
network using the bipolar and binary sigmoidal activation function.
 2. Explain the architecture of McCulloch-Pitts Neuron network model. Implement
NAND(NOT-AND) gate function using M-P Neuron Model(with binary input).

Course Outcome 2(CO2):

1. Find the weights required to perform classification of patterns shown below using
perceptron network. The patterns (1,1,-1) and ( 1, -1,-1) are belonging to the target class
-1. The patterns (-1,1,1) and (-1,-1,1) are belonging to the target class +1. Assume suitable
learning rate and initial weights.
2. Explain the architecture and training algorithm of Adaline network . Use Adaline nerwork
to train NOR logic function with bipolar inputs and targets. Perform 2 epochs of training.

Course Outcome 3(CO3):

1. There is an imprecise relationship between the ambient temperature for clay masonry bricks
and their compressive strengths. Let X be a fuzzy set of fracture strengths and Y be a fuzzy
set of temperatures with the following membership functions:

(a) Find the Cartesian Product of X and Y and represent it as relation R.

Suppose there is a second fuzzy set of masonry lengths given as

(b) Find S=ZoR using max-min composition (c) Find T=ZoR using max-product composition
 2. Given two universes X={x1,x2,x3,x4,x5} and Y={y1,y2,y3,y4,y5},the fuzzy sets A
defined on X and fuzzy set B defined on Y are given below:
0.4 0.7 1 0.8 0.6 0.2 0.6 1 0.9 0.7
𝐴=� + + + + �𝐵 = �𝑦1 + 𝑦2 + 𝑦3 + 𝑦4 + �
𝑥1 𝑥2 𝑥3 𝑥4 𝑥5 𝑦5

(i) Find the relation R = A × B

0.4 1
Consider another fuzzy set C defined on the universe V={v1,v2,v3}, 𝐶 = � + +
0.8
�
𝑣1 𝑣2 𝑣3
(ii) Find P = B × C. Using max-min composition, Find RoP.

Course Outcome 4(CO4):

1. Illustrate the various types of cross over with suitable examples.
2. Using Genetic algorithm with Roulette wheel selection method maximize the function
f(x)=x2 over {0, 1, 2, ..., 31} with initial x values of (13, 24, 8, 19). Show one crossover and
mutation.

Course Outcome 5(CO5):

1. Explain strong dominance and weak pareto-optimality.
2. What are the different classifications of neuro-fuzzy hybrid systems?

Syllabus
Module – 1 (Introduction to Soft Computing & Artificial Neural Network)
Introduction to Soft Computing. Difference between Hard Computing & Soft Computing.
Applications of Soft Computing. Artificial Neurons Vs Biological Neurons. Basic models of
artificial neural networks – Connections, Learning, Activation Functions. McCulloch and Pitts
Neuron. Hebb network.
Module – 2 (Supervised Learning Network)
Perceptron Networks– Learning rule, Training and testing algorithm. Adaptive Linear Neuron–
Architecture, Training and testing algorithm. Back propagation Network – Architecture, Training
and testing algorithm.

Module - 3 (Fuzzy Logic & Defuzzification)

Fuzzy sets – properties, operations on fuzzy set. Fuzzy membership functions, Methods of
membership value assignments – intuition, inference, Rank Ordering. Fuzzy relations– operations
on fuzzy relation. Fuzzy Propositions. Fuzzy implications. Defuzzification– Lamda cuts,
Defuzzification methods.
Module - 4 (Fuzzy Inference System & Genetic Algorithm)
Fuzzy Inference Systems - Mamdani and Sugeno types. Fuzzy Logic Controller. Concepts of genetic
algorithm. Operators in genetic algorithm - coding, selection, cross over, mutation. Stopping
condition for genetic algorithm.

Module - 5 (Multi Objective Optimization & Hybrid Systems)

Multi objective optimization problem. Principles of Multi- objective optimization, Dominance and
pareto-optimality. Optimality conditions. Neuro-fuzzy hybrid systems. Genetic – neuro hybrid
systems.

Text Books
1. [Link] and S.N. Deepa, Principles of Soft Computing , 2 ndEdition, John Wiley &
Sons.
2. Kalyanmoy Deb, Multi-objective Optimization using Evolutionary Algorithms, 1st
Edition, John Wiley & Sons.

ReferenceBooks
1. Timothy J Ross, Fuzzy Logic with Engineering Applications, John Wiley & Sons, 2016.
2. [Link], [Link] Pai “Neural Networks, Fuzzy Logic & Genetic
Algorithms Synthesis and Applications”, Prentice-Hall India.
3. Simon Haykin, “Neural Networks- A Comprehensive Foundation”, 2/e, Pearson Education.
4. Zimmermann H. J, “Fuzzy Set Theory & Its Applications”, Allied Publishers Ltd.
Model Question Paper

QP CODE:

Reg No:

Name: PAGES : 4

APJ ABDUL KALAM TECHNOLOGICAL UNIVERSITY

EIGHTH SEMESTER [Link] DEGREE EXAMINATION, MONTH & YEAR

Course Code: CST 444

Course Name: Soft Computing

Max. Marks : 100 Duration: 3 Hours

PART A

Answer All Questions. Each Question Carries 3 Marks

1. Explain the architecture of a simple Artificial Neural network? Compare it with a

biological neuron.

2. A 4-input neuron has weights 1, 2, 3 and 4. The transfer function is linear with the
constant of proportionality being equal to 2. The inputs are 4, 10, 5 and 20
respectively. Predict the output?

3. Explain the Widrow-Hoff learning rule for supervised learning in neural networks
with help of an example. Why is it sometimes called the LMS learning rule?

4. Implement one epoch of Adaline algorithm for AND logic function with binary
inputs and bipolar outputs. Initial weights are w1=0.2, w2=0.1 and learning rate
parameter η=0.2.

5. Consider two fuzzy sets 𝐴 = �0.2 + 0.3 + 1 + 0.1 + 0.5 �𝐵 = �

0.1
+
0.25 0.9
+ +
0.7
+
0 1 2 3 4 0 1 2 3

0.3 �Find the following: (a) Algebraic sum (b) Algebraic product(c) Bounded sum.
4

6. Using your own intuition and definition of universe of discourse, plot membership
 functions for liquid level (Empty, very less, less, full, very full) in a tank.

7. Explain Stochastic Universal Sampling with an example.

8. Explain any two mutation methods.

9. Differentiate between linear and nonlinear Multi Objective Optimization Problem.

10. What are the characteristics of neuro fuzzy hybrid systems?

(10x3=30)

Part B
(Answer any one question from each module. Each question carries 14 Marks)

11. (a) Implement XOR function using M-P Neuron Model (with binary input).Why (8)
M-P neuron is widely used in processing binary data?

(b) Using Hebb Network calculate the weight required to perform the following (6)
classification of given input pattern.
L belongs to the members of the class(+) target value +1
U does not belongs to members of class(.) target value -1

L U

OR

12. (a) Compare the three learning approaches in Artificial Neural Network. How is (8)
the critic information used in learning process.

(b) Define Hebb Law. Design a Hebb Network to implement logical AND (7)
function. Use bipolar input and targets.
13. (a) Discuss the training algorithm and explain the weight updates in back (10)
propagation networks.

(b) Implement one epoch of Perceptron training algorithm for OR logic function (4)
with binary input and bipolar output.

OR

14. (a) Explain how synaptic weights are adapted iteration by iteration using error (10)
correction rule in Perceptron convergence algorithm for an OR gate with
bipolar inputs and outputs. Initial weights are all zero and learning rate
parameter η=0.1.

(b) Explain Perceptron convergence theorem and discuss Perceptron algorithm (4)
based on XOR logic function.

15. (a) Three fuzzy sets are defined as follows: (10)

0.1 0.2 0.3 0.4 1 0.2 0.5 0.7 0.3 0
𝐴=� + + + }, 𝐵 =� + + + + + },
30 60 90 120 1 2 3 4 5 6

0.33 0.65 0.92 0.21

𝐶=� + + + �
100 200 300 400

Find: (i)𝑅 = 𝐴 × 𝐵 (ii)𝑆 = 𝐵 × 𝐶 (iii)𝑇 = 𝑅𝑜𝑆, using Max-Min composition

(iv)𝑇 = 𝑅𝑜𝑆, using Max-Product composition.

(b) For the fuzzy sets given 𝐴 = �0.5 + 0.2 + 0.9 and𝐵 = �1 + 0.5 + 1
}. Find (4)
}
𝑥1 𝑥2 𝑥3 𝑦1 𝑦2 𝑦3

relation R by performing Cartesian product over the given fuzzy sets.

OR

16. (a) Using inference approach, find the membership values for each of the (8)
triangular shapes (I, R, IR, T) for a triangle with angles 120°, 50°, 10°.

(b) Using Zadeh’s notation, determine the ʎ - cut sets for the given fuzzy sets: (6)
 0 0.5 0.65 0.85 1.0 1.0

𝑆1 = � + + + + + �
0 20 40 60 80 100

0 0.45 0.6 0.8 0.95 1.0

𝑆2 = � + + + + + �
0 20 40 60 80 100
 Express the following for ʎ = 0.5: a) 𝑆1 𝖴 𝑆2 b) 𝑆´2 c) 𝑆1 ∩´ 𝑆2

17. (a) Differentiate between value encoding and permutation encoding. (8)

(b) Explain the stopping conditions for genetic algorithm. (6)

OR

18. (a) Apply Mamdani fuzzy model to design a controller to determine the wash (10)
time of a domestic washing machine. Assume input is dirt and grease of the
cloth. Use three descriptors for input variable and five descriptors for output
variables .Derive the set of rules for controller action and defuzzification.
Design should be supported by figure wherever possible.

(b) Explain Single-Point Crossover and Two-Point Crossover with example. (4)

19. (a) Explain convex and non convex MOOP? How to find a non dominated set. (10)

(b) What are the properties of dominance relation? (4)

OR

20. (a) Explain Genetic Neuro-Hybrid System with block diagram. Also write the (8)
advantages of Genetic- Neuro Hybrid systems.

(b) Discuss the classification of Neuro-Fuzzy Hybrid System. (6)

 Teaching Plan

No. of
Lecture
No Contents Hours
(35 hrs)

Module-1 (Introduction to Soft Computing & Artificial Neural Network) (6 hours)

1.1 Introduction to Soft Computing 1 hour

Difference between Hard Computing & Soft Computing & Applications of

1.2 1 hour
Soft Computing

Artificial Neurons Vs Biological Neurons, Basic models of artificial neural

1.3 1 hour
networks

1.4 Activation Functions 1 hour

1.5 McCulloch and Pitts Neuron 1 hour

1.6 Hebb network 1 hour

Module-2 (Supervised Learning Network) (7 hours)

2.1 Perceptron networks – Learning rule, Training and testing algorithm 1 hour

2.2 Perceptron networks – Problems 1 hour

2.3 Adaptive Linear Neuron (Lecture I) 1 hour

2.4 Adaptive Linear Neuron (Lecture II) 1 hour

2.5 Adaptive Linear Neuron-Problems (Lecture III) 1 hour

2.6 Back propagation Network (Lecture I) 1 hour

2.7 Back propagation Network (Lecture II) 1 hour

Module-3 (Fuzzy Logic & Defuzzification) (8 hours)

3.1 Introduction to Fuzzy Set, Properties & operations on fuzzy sets 1 hour

3.2 Fuzzy membership functions, Fuzzification 1 hour

3.3 Methods of membership value assignments 1 hour

3.4 Fuzzy relations, Operations on Fuzzy Relation 1 hour

3.5 Fuzzy Propositions & Fuzzy Implications 1 hour

3.6 Lamda cuts for fuzzy sets 1 hour

3.7 Defuzzification methods(Lecture I) 1 hour

3.8 Defuzzification methods(Lecture II) 1 hour

Module-4 (Fuzzy Inference System & Genetic Algorithm) (6 hours)

4.1 Fuzzy Inference Systems - Mamdani type 1 hour

4.2 Fuzzy Inference Systems - Sugeno type 1 hour

4.3 Fuzzy Logic Controller 1 hour

4.4 Introduction to genetic algorithm, operators in genetic algorithm - coding 1 hour

4.5 Selection, Cross over 1 hour

4.6 Mutation, stopping condition for genetic algorithm 1 hour

Module-5 (Multi-Objective Optimization & Hybrid System) (8 hours)

5.1 MOOP-Linear &Non linear, Convex & Non Convex 1 hour

Principles of MOO-Illustrating Pareto Optimal Solutions, Objectives in

5.2 1 hour
MOO

5.3 Dominance & Pareto-Optimality-Concept of Domination 1 hour

5.4 Properties of Dominance Relation, Pareto Optimality 1 hour

5.5 Procedure for finding a non dominated set 1 hour

5.6 Optimality Conditions 1 hour

5.7 Neuro Fuzzy hybrid system-Classification& characteristics 1 hour

5.8 Genetic –neuro hybrid systems 1 hour