Special Topics in Computational

Intelligence
Dr. Muhammad Umair
Assistant Professor
Faculty of Information Technology
University of Central Punjab

1
Lecture No. 01
• Contents
• Self Introduction
• Marks Distribution
• Discussion on course contents
• Course Reference Books
• Expectation by the end of semester
• What is Computational Intelligence?
• Discussion between biological structures and CI tools
• Application of different areas of CI.
• Outline of any research paper.
• How to develop your interest in any research topic.
• Types of research papers
• Difference between conference and journal paper
• Important CI conferences and journals

2
Marks Distribution
• 15% Mid Tem Exam
• 30% Final Exam
• 15% Quizzes
• 40% Research Paper Presentation
• 10% Critical review
• 10% implementation
• 10% Improvement suggestion and implementation
• 10% Papers Submission

3
Course Contents

4
 Week 1-8
Lecture

Metho
Assess
Week

ment
Topic
No.

d
1 1-2  Introduction to the course contents, How to use them in learning process?
 Overview literature regarding the fundamental ideas and researchers of Computational Intelligence.
 Discussion of writing first research paper and choosing research topic.
 Discussion on Key Computational Intelligence conferences, journals and societies.

2-4 3-8 EVOLUTIONARY COMPUTING: Quiz 1

Introduction to Evolutionary Computing, Representation of Solutions - The Chromosome, Fitness Function, Initial
Population, Selection Operators,
Types of Selection Operators:
Random Selection, Proportional Selection, Tournament Selection, Rank-Based Selection, Elitism.
5-6 9-12 -Genetic Algorithm. Paper Presentations
-General Concept of GA.
-Different types of GA. Quiz 2
-Different Strategies for using GA concept.
-Discussion on Literature relevant to GA.
-Application of GA in real world.
-Project GA.
-Student Paper presentations
7-8 13-16 -Swarm Optimization and Ant Colony Optimization Quiz 3
-Discussion on Several important literatures of swarm optimization
-General concept and features of swarm optimization Paper Presentations
-Different types of swarm optimization algorithms
-Application of swarm optimization in real world.
-Project PSO.
5
- Student Paper presentations
Week 10-16
10-11 19-22 Artificial Neural Networks Quiz 4
-Important literatures of ANN
-General concept of ANN Paper Presentations
-Different types of ANN (Supervised and Unsupervised)
-Application of ANN in real world
-Project ANN
-Convolutional Neural Network
Student Paper presentations

12-13 23-26 Fuzzy Logic Quiz 5

-Important literature of on Fuzzy systems.
-General concept and features of fuzzy systems. Paper Presentations
-Different strategies of using Fuzzy systems.
-Application of Fuzzy systems
-Project FL
Student Paper presentations

14 27-28 -Discussion on Papers

15-16 29-32 -Student Paper presentations Paper Presentations

6
Textbook: Computational Intelligence: Concepts to Implementations
by Eberhart and Shi

7
Computational Intelligence, An Introduction
By
Andries P. Engelbrecht

8
Expectation by the end of semester
• Grip on understanding implementing
• Modern Computational Intelligence concepts
• One conference paper in any application area of CI.

9
What is Computational
Intelligence?

10
Computational Intelligence Definition

Computational intelligence comprises practical adaptation and self-

organization concepts, paradigms, algorithms and implementations
that enable or facilitate appropriate actions (intelligent behavior) in
complex and changing environments.

11
Artificial Neural Network

12
Biological Neuron

13
Biological Basis: Neural Networks
Neurons: nerve cells; consist of dendrites, body and
an axon; signals flow through synapses.

Some differences between biological and artificial

neurons (processing elements):
* Signs of weights (+ or -)
* Signals are AC in neurons, DC in Pes
* Many types of neurons in a system; usually only
a few at most in neural networks
* Basic cycle time for PC (~100 ns) faster than
brain (10-100ms) {as far as we know!}

14
Application Areas: Neural Networks

• Classification
• Which of the best classes defines an input parameter?
• Content Addressable Memory
• A complete version obtaining at output of network by providing partial
version of input.
• Compression
• Compression and decompression.
• Composition
• Generation of keys and patterns

15
Ctd…
• Control Systems
• ECG waveform classification.
• Future market trends prediction.
• Stock exchange trends prediction.
• Targets tracking / Recognition
• Text to speech conversion
https://www.youtube.com/watch
?v=qv6UVOQ0F44
Evolutionary Computing

18
Application Areas: Evolutionary Computation
• Optimization
• Fault diagnosis.
• Robot track determination.
• Schedule Optimization.
• Load distribution by an electric utility.
• Product ingredients optimization.
• Classification.

19
Fuzziness in life

20
Application Areas: Fuzzy Logic
•Control systems
* Vehicles
* Home appliances
•Expert systems
* Industrial processes
* Diagnostics
* Finance
* Robotics and manufacturing

21
The possible outline of any research paper
• Abstract
• Keywords
• Introduction
• System Model/Methodology
• Results Summary
• References
How to develop your interest?
• I have shown you there graphical views till now?
• Do you got any interest?
• If still not, then…
• Try to find put papers relevant to this area
• Researcher android app
• Figure out the papers
• Other ways
• How to scrutinize the paper?
• Abstract and summary reading philosophy
• List down the papers that are understandable
Ctd…
• How to finalize develop the interest in paper?
• Read the introduction part carefully for getting the idea that
• What this contribution have relationship to previous research gaps?
• Try to do so with all papers.
• Now you will be at stage of finalizing the paper you are going to work for?
• For this purpose read the system model/methodology section
• Model that is used in paper to test the idea
• What is positivity in paper?

• This whole practice will help you to find out the research gap that can
be your future Conference and journal paper?
Types of Research articles
• Research Articles
• present fully developed analysis, results, and discussion
• Research Letters
• on major scientific advances in all major geoscience disciplines
• Commentaries
• provide reflections or context around a recent article
• Frontier Articles
• are invited articles that present a review of recent cutting-edge advances in a
leading scientific field
Difference between conference and journal
paper
• Conference papers cannot be republished without substantial
additional technical material.
• A conference is where you actually turn up to hear a paper presented as a
talk. The paper is later printed in the proceedings.

• Journal papers refer to an research article that's published in an issue

of the journal
• Can be a improved conference paper.
Important CI conferences
• CVPR : IEEE Conference on Computer Vision and Pattern
Recognition, CVPR
• NIPS : Neural Information Processing Systems (NIPS)
• International Conference on Machine Learning (ICML)
• CM SIGKDD International Conference on Knowledge discovery and
data mining
• Meeting of the Association for Computational Linguistics (ACL)
• EDAS
• http://www.guide2research.com/topconf/
Top CI Journals
• https://www.scimagojr.com/journalrank.php?category=1702
• Way to check the journal originality
• http://mjl.clarivate.com/
• Finding out the impact factor of journal
• https://guides.hsl.virginia.edu/faq-jcr
I hope you enjoyed lecture
• My contacts:
• 03335610677
• 042-35880007 ext 464
• [email protected]

• Fb: umairbwp

• I am in search of hardworking and motivated students for continuing

their research with me.

29
Thankyou

30
Setting the base of Computational
Intelligence
• Adaptation

• Self Organization

• Literature about Computational Intelligence

• How it fits in to models of Intelligent Systems

Adaptation
• Ability to change its parameters in order to better meet its goals
• Dynamic Adaptation
• Real time adaptation
• Not offline adaptation
• System adapts while its running
Difficulties in Dynamic Adaptation

• A large problem space which may contain many solutions

• A large number of variables in each direction

• Function measuring performance is complex and non linear

• Have local optima and discontinuities

• Fitness function varies with time

• A complex and changing environment in which it exists