COSC 6367 --- Evolutionary Programming
( Dr. Eick )
| The Joy of Living - Max Ernst (1936) |
2012 Course Summary
Evolutionary Computing studies a wide range of problem-solving
techniques which are based on principles of biological
evolution, such as natural
selection and genetic inheritance. These techniques are currently
widely applied to a variety of problems, ranging from practical
applications in industry and commerce to leading-edge scientific research.
The course will be giving an introduction to evolutionary computing and
discuss the application of evolutionary computing to search, optimization,
machine learning, design, simulation of evolution in biological and
other systems, and art.
It will also provide a
sound introduction to related fields, namely search and numerical optimization.
Moreover, we will take
a closer look at approaches which try to understand what rules guide the
evolution of complex systems, such as the evolution of cities, and
at work which centers on developing agent-based systems which solve
problems collaboratively. In particular, in the part of the course
agent-based modeling and simulation, software platforms for agent-based
modelling, and swarm intellgence approaches
will be covered.
The course is quite "hands-on" and project-oriented and with less emphasis on
exams; there will be
will be a final exam and two review-style quizes. As the course covers a
smaller
field more time will be allocated to explaining principles
and applications of evolutionary computing in more depth, and in assisting students with the course projects. There will be 3 course projects, including a
group project which count 50% towards the course grade. The course exam
and the two quizzes ount 49% towards the course grade, and 1% are
allocated to attendance and extra credit. Overall, the workload for the course
will be less than for the Data Mining course I teach every fall semester (about 75%, is my best guess).
News COSC 6367
- I enjoyed teaching the course and wish you an interesting summer!
- Your letter grades for the course should now be visible to you.
Zechun will e-mail you a detailed grade report early next week.
The final exam will not be returned to students;
however, you can see you final exam it the following dates:
May 16, 11a-noon, August 30, noon-1p, September 4, 2-3p.
- These are the 2012 weights of the different parts of the course
that were used when computing the final grade:
Project 1: 23%,
Project 2: 17%,
Project 3: 11%,
Quiz 1: 11%,
Quiz 2: 12%,
Final Exam: 25%,
Attendance: 1%.
- Project1 scores are raw scores and still subject to curving.
- Tentative Teaching Schedule for the Remainder of the Semester: April 3: Spatial Cons.; April 5:
Evolutionary Art; April 10: Quiz2; April 12: Theory of EC; April 17: LCS.
April 19: Project2 Student Presentations and Demos; April 24: LCS & XCS
April 26: Leftovers and Review for Final Exam; May 8: Final Exam.
Course Information Spring 2012
Syllabus:
Link to Spring
2012 COSC 6367 Syllabus
Prerequisites: the course is more or less self-contained, students should have sound programming skills;
students can use programming languages of their own preferences in course project.
Focus: Theory and application of evolutionary programming and other related areas in evolutionary and natural
computation centering on genetic algorithms and programming, evolution strategies,
classifier systems, artificial life, and other models that rely on evolutionary principles. Students will perform
course projects that apply the discussed techniques to optimization problems, to develop adaptive systems, to the
simulation of biological and cultural systems, and to producing computer art.
Teaching Philosophy: The course will be very project oriented; you will develop medium-sized software systems
that employ evolutionary computing paradigms in 3 course projects. Alternatively, you can choose to
investigate a subject and give a oral presentation (+ report or webpage) as your third project.
In addition to regular lectures there will be a
lot of project related discussions during the lectures.
Textbook: A.E Eiben and J.E. Smith, Introduction to Evolutionary Computing, Springer, second edition or later, 2007 or later
(Eiben Textbook Transparencies)
The 6367 Team
Instructor: Dr. Christoph F. Eick
Office hours: TU 2-3:15p and TH 1-1:45p in 589 PGH
e-mail: ceick@uh.edu
TA: Zechun Cao
Office hours: TU/TH 9:30-10:30a (in 313 PGH)
Zechun's
6367 Website
TA Email: zechun.eps12@gmail.com
6367 Google News Group
Tentative 2012 Course Schedule
Quiz1: Th., March 8
Quiz2: Tu. April 10
Final Exam: Tu., May 8, 11a(!)-1:10p
Deadlines:
Project 1: March 1
Project 2: Short report on April 18, 15-25 minute demos on April 19
Project 3: Fr., May 4, 11p (submit your report to both Zechun and
Dr. Eick
2012 Reviews
Feb. 28 Review1 Questions and Answers
Review2 on March 6 will go through the search algorithms introducted earlier
April 5 Review3 (will mostly discuss the Swarm Intelligence Book Chapter)
May 1 Review4 (will go through some older final exam questions and will
discuss the solutions to Quiz2)
2012 Class Transparencies
Course Organization and Introduction
Evolutionary Computing:
Course Information (Part1,Part2)
Knapsack Problem
(will be used as an example for how EC works).
Obitko.com
Tutorial (will be used to demo some EC systems)
Example Applying EC to the Travelling Salesman Problem
Survey Evolutionary Computing:
Eiben Chapter1
Eiben Chapter2
Student February 2010 Questions
Eiben Chapter3 Genetic Algorithms
Overview ES
Eiben Chapter4 Evolution Strategies
Eick on Using EC for Transportation Problems
Project1-07
Eiben Chapter6 Genetic Programming
2003 Koza & Poly Genetic Programming Tutorial
A Field Guide to
Genetic Programming
Review of Popular Search Techniques (Search1,
Search2,
Search3)
What is Unique about EC---if
Compared with Other Search Teachniques?
Eiben Chapter8 Parameter Control
On Numerical Optimization Problems
Constraints Handling in Numerical
Optimzation (Tutorial by Cuello; added in 2012; slides 1-28, 38-41, 49-51,
59 and 79-81 will be discussed in the lecture)
Swarm Intelligence/Agent-based Modeling:
Book Chapter on Swarm Intelligence
by D. Corne, A. Reynolds, and Eric Bonabeau (please read the
whole chapter, except you can skip Section 3.1
as Ant Colony Optimization will not be discussed in COSC 6367; moreover,
there are slides associated with that article: Dr. Eick's Discussion of the
Book Chapter and Particle Swarm Optimization (PSO)---which
might be worth to look at)
Overview what will be covered about
Swarm Intelligence and Agent-based Modeling
Swarm Intelligence: Nature's way to system
engineering (We will scan through first 25 slides of this excellent
introductory tutorial on Swarm Intelligence by Gianni Di Cara from the
University of Lugano in Switzerland)
Wikipedia on Agent-based Models
Macal&North's Introduction to Agent-based Modeling and Simulation
(just look at slides 1-10 and 24-30)
Drogoul's Introduction to Agent-based Systems (slightly
better and more up-to-date than the Macal/North-Introduction; slides 5, 11-12, 19-22, 27-34, 37, and
49-52 will be covered in the lecture on April 3, 2012)
Introduction NetLogo Agent-based Modeling
Package
Swarm Robots (Swarmanoids are swarms of interacting, diverse robots; a
Smarmanoids video can be found at
Marco Dorigo's Website,
Havard's Kilobot Website---Kilobots a very small and cheap swarm robots and
Robots which fly and collaborate---watch demonstration which
starts at 10:43)
Introduction to Swarm
Intelligence and its Application to Optimization (subset of
slides (numbers 1-25, 38, 47-50, 63) of a tutorial of Li and Engelbrecht are used in the lecture)
More on Evolutionary Computing:
Eiben Chapter9 (Spatial Considerations)
Eiben Chapter13 (Evolutionary Art)
Eiben Chapter11 Theoretical Foundation of EC
Co-Evolution (might or might not be discussed based on time)
Eiben Chapter10 (Memetic Alg.; might not be
covered in 2012)
Evolutionary Learning:
Dr. Eick's Introduction to Machine Learning with
EC
XCS
Tutorial by Stewart Wilson (in pdf-format)
Bull Paper Review
Introduction to Reinforcement Learning
Probabilistic Model Building Genetic Algorithms
Learning to Predict and Classify (
added on April 18, 2010)
Miscellaneous:
Introduction to Artificial Life
Brief Introduction to Soft Computing (only
the first 11 transparencies will be used)
Will Brown's Presentation on Useful AI Techniques
Teaching Material of the Spring 2001 Teaching of
Evolutionary Programming
2012 Course Projects
Project1
Project2 (Deliverables and Deadlines: A short report (discussing
1. Project2 Goals 2. Approach Chosen 3. Summary) will be
due on April 18 and groups will survey and demo their work during
the class on April 19) will be a group project and
will focus on emergent behavior, swarm intelligence, and agent-based modeling
and simulation in
general and on simulating the
behavior of wolf
packs (also see A Day of a Wolf in a Wolf Pack,
WolfQuest Game,
Example Wolf-Sheep-Grass Simulation,
Swarm Behavior) in
particular. Activities will start after
Spring break, but it would be great, if you could do little reading during
Spring break. Likely, we will use REPAST in the project, although groups
might also consider NetLogo
to implement their agent-based modeling and simulation systems; NetLogo models
can be imported into REPAST.
Project 3
Grading
Students will be responsible for material covered in the
lectures and assigned in the readings. All assignment and
project reports are due at the date specified.
No late submissions
will be accepted after
the due date. This policy will be strictly enforced.
Seveal times during the semester I will check class attendance at randomly
chosen dates, and an attendence score will be computed from how many
of the those lectures you attended.
Translation number to letter grades:
A:100-90 A-:90-86 B+:86-82 B:82-77 B-:77-74 C+:74-70
C: 70-66 C-:66-62 D+:62-58 D:58-54 D-:54-50 F: 50-0
Only machine written solutions
are accepted (the only exception to this point are figures and complex formulas) in the assignments.
Be aware of the fact that our
only source of information is what you have turned in. If we are not capable to understand your
solution, you will receive a low score.
Moreover, students should not throw away returned assignments or tests.
Students may discuss course material and homeworks, but must take special
care to discern the difference between collaborating in order to increase
understanding of course materials and collaborating on the homework /
course project
itself. We encourage students to help each other understand course
material to clarify the meaning of homework problems or to discuss
problem-solving strategies, but it is not permissible for one
student to help or be helped by another student in working through
assignment problems and in the course project. If, in discussing course materials and problems,
students believe that their like-mindedness from such discussions could be
construed as collaboration on their assignments, students must cite each
other, briefly explaining the extent of their collaboration. Any
assistance that is not given proper citation may be considered a violation
of the Honor Code, and might result in obtaining a grade of F
in the course, and in further prosecution.
Useful Links Spring 2010
Bull's "Introduction to Learning
Classfier Systems;
Butz' XCS Software
Hitchhiker's Guide to Evolutionary Computing
Biologically Insprired Robotics Network (Biro-net aims to understand the underlying mechanisms that allow natural and robotic agents to adapt and survive in uncertain and dynamic environments).
DEMO (Dynamic and Evolutionary Machine Organization)
On Genetic Drift
Reinforcement Learning Repository at UM Amherst (reinforcement
learning and classifier systems have a lot in common)
Epimorph Virtual Creatures(Video,
general information)---donated by Paul Ledbetter III.
Image Creation through
Cellular Automata "Ying Yang Fire"
Previous Quizzes and Exams
Quiz1 2012 (Answer Sheet for Quiz1)
Quiz2 2012 (Solution Sketches for Quiz2)
Quiz2 2010 (Quiz2 in 2010)
Quiz1 2008 (Answer Sheet for Quiz1)
Quiz2 2008 (Answer Sheet for Quiz2)
Quiz1 2007 (Answer Sheet for Quiz1)
Quiz2 2007 (Answer Sheet for Quiz2)
2010 Final Exam with Solution Sketches
2007 Final Exam
2008 Final Exam
Solution Sketches 2012 Final Exam
EC Software
Jing Zhang's
Survey EC Programming
Environments
Java Genetic Algorithm Library---you are allowed
to use this library or any other library for the course programming projects---donated by Abraham Bagherjeiran.
ECJ---a Java-based EC Research systems (GMU)
Discipulus Genetic-Programming Software
Sipper's Genetic Programming
for Games
LISP-Code of
a "simple" Symbolic Regression System that employs Genetic Programming.
To run the system save the file as gp.l, login into a Sun-workstation
and type:
akcl
(load "gp.l")
(run-gp 30 #'eval1)
(bye)
Journals, Conferences, and Books
Evolutionary Computation Journal
2007 Projects
Project1: Project1 Specification,
Supplement
to the Specification of Project1. Moreover, Paul Ledbetter used an
optimization package to obtain the following "optimal"
solutions for cost-function1 --- the linear cost function. Solutions for the Quadratic
Transporatation Problem (generated by Justin Thomas).
Project2:
Specification
Project3: Specification
2010 Projects
Project1: Solving Balanced Transportation
Problems (What to submit no later than
March 6, 11p to Chun-sheng)
Project2: Survey on a Subfield of Evolutionary Computing (group project;
each group will investigate a different theme)
Project3: Learning to Predict with
Genetic Programming
2008 Projects
Project1: Travelling Salesman
Project1 Specification
TSP Benchmark
Project2: Developing Adaptive Systems
Project3 2008
last updated: May 13, 2012
created by YingYangFire
