UNIVERSITY INSTITUTE OF ENGINEERING

DEPARTMENT OF AIT - CSE

Bachelor of Engineering (CSE)
SOFT COMPUTING (22CSH-345 )
By: Ms. Aarti (E15380)

DISCOVER . LEARN . EMPOWER

 Course Objectives
To introduce soft computing concepts and techniques of artificial neural networks,
fuzzy sets, fuzzy logic and genetic algorithms

To understand the various techniques from the application point of view.

To analyze various soft computing techniques and decide the technique to be used in
a particular problem situation.

To implement soft computing based solutions for real-world problems

2
 Course Outcomes
CO Identify and describe soft computing techniques and 1
1 their roles in building intelligent. Machines
CO Recognize the feasibility of applying a soft 2,4
2 computing methodology for a particular problem.
CO Apply fuzzy logic and reasoning to handle 3
3 uncertainty and solve engineering problems, genetic
algorithms to combinatorial optimization problems
and neural networks to pattern classification and
regression problems.
CO Effectively use modern software tools to solve real 3
4 problems using a soft computing approach.
CO Evaluate various soft computing approaches for a 4 3
Table of Contents
• Introduction to soft computing
• Difference Between Soft computing and Hard Computing
• Major Areas of Soft Computing
• Applications of soft computing

4
 Introduction to Soft Computing

Soft Computing is the collection of computational techniques in Computer Science,

AI, Machine learning and some engineering disciplines which attempt to study, model
and analyze very complex phenomenon – those for which conventional methods have
not yielded lost cost, analytic and complete solutions.

Some of it’s principle components includes:

• Neural Network(NN)
• Fuzzy Logic(FL)
• Genetic Algorithm(GA)

5
Concept of Computation

6
Properties of Computing
• Should provide precise sol.
• Control action should be unambiguous and accurate
• Suitable for problems that are easy to model mathematically.

7
Hard Computing
• LAZ (Lotfi Aliasker Zadeh) in 1996.
• As per him, hard computing gives
• Precise results
• Steps are unambiguous
• Control action is formally defined by a mathematical model/algo.

8
Examples of Hard
Computing
• Solving numerical problems such as roots of a polynomial, integration,
differentiation etc.

• Searching and sorting algorithms give precise results with defined

algo.

• Shortest tour in a graph.

9
 SOFT COMPUTING
(SC)
SC was coined by LAZ.

Lotfi A. Zadeh (Inventor of fuzzy logic) discovered soft computing. He

describes it as follows:

“Soft computing is a collection of methodologies that aim to exploit the

tolerance of imprecision and uncertainty to achieve tractability, robustness &
low solution cost. “

Role model of SC is human brain.

Characteristics of soft computing
• It does not require any mathematical modelling of problem
solving.
• It may not yield precise solution.
• Algorithms are adaptive in nature.
• Use some biological inspired methodologies such as
genetics, evolution etc.

11
Components of soft computing
•Fuzzy Logic (FL),
•Artificial Neural Networks (ANN),
•Evolutionary Computation (EC),
•Swarm Intelligence (i.e. Ant colony optimization and Particle swarm
optimization, )
•Additionally Some Machine Learning (ML) and Probabilistic Reasoning
(PR) areas

12
 Hard c omp ut i ng v s Soft computing
Hard Computing Soft c o m p u t i n g
Precisely stated analytical model Imprecision is tolerable
required
More Computation time required A s it involves intelligent
computational steps,
computational time required is less
It involves binary logic crisp systems It involves nature inspired systems
and numerical analysis such as neural networks, fuzzy
logic systems and swarm
intelligent system.
Precision is observed within the Approximation is obtained
computation
in the computation
Imprecision and uncertainity are Tolerance for imprecision and
undesirable properties uncertainty is exploited to achieve
tractability, lower cost, high Machine
intelligence quotient and economy
of communication.
It produces precise answers It can produce approximate answers
 Hard computing v s Soft
computing
Hard Computing Soft c o m p u t i n g
Programs are written which follow Programs are evolved which
standard rules of programming require new laws and theories to
be created and justifi ed while
programming
The outcome is deterministic(i.e., The outcome is stochastic or
Every trial run, the output is same) random in nature and need not
be deterministic
It requires exact input data It can deal with ambiguous and
noisy input data
It strictly follows sequential It allows parallel computations
computations
 TEXTBOOKS / REFERENCE
BOOKS
TEXT BOOKS
• T1. Timothy J. Ross, “Fuzzy Logic with Engineering Applications”.
• T2. S.N.Sivanandam, S.N Deepa, “Principles of Soft Computing”
• T3. Lofti Zadeh “Fuzzy Logic and Soft Computing” Word Scientific, 1995.
• T4 Samir Roy,Udit Chakraborty, “Introduction to Soft Computing: Neuro-Fuzzy and Genetic Algorithms”,
Pearson.

REFERENCE BOOKS
• R1. Bart Kosko, “Neural Network and Fuzzy Systems: A Dynamic System Approach to Machine” Prentice-
Hall 1998
• R2. L. Fausett, “Fundamentals of Neural Networks: Architectures, Algorithms, and Applications”,
Prentice-Hall, 1994.
• R3. Jack M. Zurada, “Introduction to Artificial Neural Systems”, PWS Publishing Co., Boston, 2000.
• R4 J S R Jang, “Neuro-Fuzzy & Soft Computing,”, Pearson.
15
 References

Online Video Link

• https://onlinecourses.nptel.ac.in/noc22_cs54/preview
• https://archive.nptel.ac.in/courses/106/105/106105173/
• Text Book
• Bart Kosko, “Neural Network and Fuzzy Systems: A Dynamic System Approach to Machine” Prentice-Hall 1998

• L. Fausett, “Fundamentals of Neural Networks: Architectures, Algorithms, and Applications”, Prentice-Hall, 1994.

• Jack M. Zurada, “Introduction to Artificial Neural Systems”, PWS Publishing Co., Boston, 2000.

• J S R Jang, “Neuro-Fuzzy & Soft Computing,”, Pearson.

16
THANK YOU

For Queries:
aarti.e15380@cumail.in

17