Prerequisites: CS 302, MATH 330
Syllabus
CS 457: Computational Intelligence
Winter 2008
|
Meeting Times
|
Lect: 12:00 -
12:50 MWF, HB 112
Labs:
12:00 - 12:50 Tuesday, HB 204
|
|
Instructor
|
Dr.
Razvan Andonie, HB 219-B, Office hours
|
|
TA
|
Lukas
Magill, MagillL@gwmail.cwu.edu
|
|
Texts
|
Russell C.E., Yuhui S. Computational Intelligence Concepts to
Implementations, Elsevier, 2007.
|
Objectives
Introducing
concepts, models, algorithms, and tools for development of intelligent systems.
Example topics include artificial neural networks, genetic algorithms, fuzzy systems, swarm
intelligence, ant colony optimization, artificial life, and hybridizations of
the above techniques. This domain is called Computational Intelligence, and it
is a numerical interpretation of biological intelligence.
Topics:
Mimicking
Nature for Problem Solving: The Basic Concepts
Single-Layer and
Multi-Layer Feedforward Neural Networks
Feedback Neural
Networks
Associative
Memories
Learning Vector Quantizer
Self-Organizing
Feature Maps
Radial Basis
Function Neural Networks
Genetic
Algorithms
Fuzzy Sets and
Fuzzy Logic
Fuzzy Neural
Networks
Fuzzy ARTMAP
Swarm Intelligence
and Ant Colony Optimization
Artificial Life
The Power and
Computational Complexity of Computational Intelligence Models
Student Learning Outcomes
On
completion of this course, the student will have:
-
An understanding of the fundamental
Computational Intelligence models
-
Implemented neural networks, genetic
algorithms, fuzzy neural networks, and ant colony optimization algorithms.
-
Applied Computational Intelligence
techniques to classification, pattern recognition, prediction, rule
extraction, and optimization problems.
Grading
The will be no
exams. Grading will be based only on attendance, assignments, in-class
presentations, and the Final Project:
10% Class Participation
40% Assignments and Presentations
50% Final Project
The
final project is to be completed in groups of two, or individually. It
should be submitted as a report and presented in class. A
project progress report (1/2 page) should be submitted periodically (see Course
Schedule). The
project should contain the problem description, the code, the benchmarking, and
a discussion. For a good presentation, read Quinn's effective
presentation techniques guidelines.
Grade Distribution:
95 - 100 A
90 - 94 A-
87 - 89 B+
83 - 86 B
80 - 82 B-
77 - 79 C+
73 - 76 C
70 - 72 C-
67 - 69 D+
63 - 66 D
60 - 62 D-
0 -
59 F
Current Final Projects
NOTE:
The projects may be continued: CS496 Individual Study (up to 4 credits) or
CS458 AI Project (2 credits).
Final Projects From
Previous Years
2007/08
Ant Colony Optimization for Graph
Coloring
-
Brandon Wysocki, Benjamin Sisson
Ordered
Fuzzy ARTMAP for Predicting HIV-1 Protease Inhibitors - Christopher B.
Abdul-Wahid, Grant I. Barker
Genetic algorithms in investment
decision making modeling - Michael
Wilson, Steven Curtis
Attack
detection in computer networks using a neural network approach -
Hsun-Hung Chien
A Genetic Approach for Crowd Simulation
-
Kevin
Venables, Jason White
2006/07
Data
Clustering with Particle Swarms - Jarrett Sauby
SexualGA:
Gender-Specific Selection for Genetic Algorithms - Robb Berry
User Authentication using Backpropagation - Ted Steffen
Computational Intelligence Techniques
for Solving Rubik’s Cube - Joel Giese, Berk
Erkul
Modeling Civil Violence: An Evolutionary
Multi-Agent Approach - Charles Stahl, Abe Dybvig
A Neural Network Application for Attack
Detection in Computer Networks - Toby Lewandoski
A Self-Organizing Map Approach for
Clustering of XML Documents - Sorin Bucse
2005/06
Web Data
Mining -
Jessica and Aaron Berglund
Fuzzy ARTMAP Function Learning for Chess - Luke Magill
Ant Colony Optimization Applied to the Seega Game - Andrew Szydlowski,
Andrew Howell
CWU Monsters Game - Luke Arntson, Loren Harris
Code Breaking - Chris Lee
Language Processor - Blaire Sherman
Password Authentication using Backpropagation - Jonathan Widger
2004/05
Computational
Chemistry
- Sarah Abdul-Wahid, Catharine Collar
Voice
Recognition
- Virgil Mednick
Email Data
Mining
-
Robert Mitchell
Swarm Theory and
Enemy Armor
- Eric Newhouse
Optimization of
Buffer Sizes in Asynchronous Assembly - Yu-chih Hsiao
2003/04
Music
Notation Convertor: Andreas Stefik
Swarm
Intelligence in Checker: Kelly McBride, Tye
Wick
Neural Networks
in Chess:
Ashur Odah, Joe Lemley, Phil Leder, Pushpinder Heer
Spam Filtering
using Computational Intelligence Techniques: Aaron St. John,
Cory Stuart, Michael West
Japanese Kanji
Character Recognition: Andrew Pickard, Benjamin Rooney
Customer
Satisfaction Measuring using Neural Networks: Logan Riggs,
Tom Rockey, Michael Pankiewcz
Lectures
The
slides for lectures and additional material can be found in the shared
directory on Neve.
Software
The
source code of the textbook is written in C and C++ and is shareware. It can be
downloaded from http://www.computelligence.org/issue/CICI/CICI.html.
You are welcome to use it for classroom or personal learning. Here is the list
of the software modules:
Genetic algorithm
Particle swarm optimization
Back-propagation neural network
Learning vector quantization neural
network
Self-organizing feature map neural
network
Evolutionary back-propagation neural
network
Fuzzy rule system
Evolutionary fuzzy rule system
Fuzzy evolutionary fuzzy rule system
Neural network explanation facility
Free
software can be also found at:
Stuttgart Neural Network Simulator
The Java Object Oriented
Neural Network (JOONE) is an open source project that offers a highly
adaptable neural network for Java programmers. The source code is freely
available. JOONE can allow you to create neural networks easily from a Java
program. JOONE supports many features, such as multithreading and distributed
processing. This means that JOONE can take advantage of multiprocessor
computers and multiple computers to distribute the processing load.
ECG 11: A Java-based Evolutionary Computation and Genetic Programming Research System
(George
Mason University).
Additional
Reading
-
Zurada, J. Introduction
to Artificial Neural Systems. West Publishing Company, St. Paul, 1992. I
have used this textbook for several years in Brasov (Romania).
-
Haykin, S. Neural
Networks - A Comprehensive Foundation. Macmillan College Publishing
Company, New York, 1999 (second edition). It is generally considered as one
of the best.
-
de Castro, L.N.
Fundamentals
of Natural Computing. Chapman & Hall/CRC, 2006. Presents a
wide-ranging survey of techniques of nature-inspired computing.
-
Engelbrecht, A.P. Fundamentals
of Computational Swarm Intelligence, Wiley, 2005. A good overview of
evolutionary computation, particle swarm optimization (PSO), and ant
algorithms, with their connections to optimization theory.
-
Engelbrecht, A.P.
Computational Intelligence: An
Introduction, Wiley, NY, 2007, second edition.
Arguably, the best compendium on computational intelligence, but not a good
textbook.
-
Jones,
M. T.
AI
Application Programming, second edition, Charles River Media,
Hingham, Massachusetts, 2005. It has some useful implementations.
Links
Gurney's Notes help you to
learn about the basic concepts and ideas in artificial neural networks.
DSI Neural Networks Group
is
a complete source of electronic information
Genetic Algorithm
Tutorial
- this site gives a short introduction to
genetic algorithms.
Machine Learning Datasets
is
a very good source of benchmarks for testing your neural implementations.
Ant Colony Optimization
is
a source of tutorials.
SwarmWiki, hosted by the
Center for the Study of Complex Systems at the University of Michigan.
Particle swarm
optimization (PSO) is a population based stochastic optimization
technique developed by Dr. Eberhart and Dr. Kennedy
in 1995, inspired by social behavior of bird flocking or fish schooling.
A Brief History of
Connectionism by D.
A. Medler, Neural Computing Surveys, 1, 1999, 61-101.
IEEE
Computational
Intelligence Society
Neural Computing Research Programs
Worldwide
Fuzzy Sets and Systems,
a nice collection of resources from Robert Fuller.
Neural Java: Neural Networks Tutorial
with Java Applets is a series
of exercises and demos. Each exercise consists of a short introduction, a
small demonstration program written in Java (Java Applet), and a series of
questions which are intended as an invitation to play with the programs and
explore the possibilities of different algorithms. You can download the
source code of the applets.
Multiagents, contains many links on
intelligent agents.