First Things First This is 4003-590-02 / 4005-756-02 Welcome to - - PDF document

1

Welcome to Genetic Algorithms

First Things First

 This is 4003-590-02 / 4005-756-02

(Genetic Algorithms)

 I am Joe Geigel…your host!

Plan for this afternoon

 Logistics  What is this course about?  Requirements and Deliverables

First some thanks

 To Prof Butler for filling in for me.  To WGBH for making that engrossing

video on evolution.

 Time for attendance.

Logistics

 Course Web Site:

 http://www.cs.rit.edu/~jmg/geneticAlgorithms

 Everything you need to know

 Syllabus  Assignments  Schedules  Reading List  Diary

Logistics

 Course Web Site:

 http://www.cs.rit.edu/~jmg/geneticAlgorithms

 Contact:

 office hours: MW 2-4 (or by appt)  Office: 70 (GCCIS) Rm 3527  e-mail: jmg@cs.rit.edu  phone: 475-2051

 Slides:

 Will be available on Web site.

2

Logistics

 Course Web Site:

 http://www.cs.rit.edu/~jmg/geneticAlgorithms

 Will be the authoritative source for the

course.

 E.g. SCHEDULE has already been

changed.

Logistics

 Your RIT e-mail

 Be sure that it is forwarded correctly.

 myCourses

 Email  dropboxes

Logistics

 Textbook

 Evolutionary Computation by

DeJong

 Selected papers from GA

literature (on Web site)

 Supplement lectures

 Not necessary, but nice to

have as references.

What is this course about? Top 5 misconceptions about this course.

1.

Only cover GAs

2.

Will have an awfully hard final exam

3.

Hey, aren’t you Prof Anderson?

4.

Will cover an awful lot of theory (leading to an awful final exam)

5.

Long PBS documentaries will be shown each class.

Evolutionary Algorithms

 An EA uses some mechanisms inspired by biological

evolution: reproduction, mutation, recombination, natural selection and survival of the fittest.

 Candidate solutions to the optimization problem play

the role of individuals in a population, and the cost function determines the environment within which the solutions "live".

 Evolution of the population then takes place after the

repeated application of the above operators.

3

Course objectives

 Upon completion of this course, students will

be able to:

 Explain the of the principles underlying

Evolutionary Computation in general and Genetic Algorithms in particular.

 Apply Evolutionary Computation Methods to find

solutions to complex problems

 Analyze and experiment with parameter choices in

the use of Evolutionary Computation

 Summarize current research in Genetic Algorithms

and Evolutionary Computing (GRAD Only)

Getting back to the video

 Things that you unexpectedly got from the

video:

 Monkeys are capable of learning the concept of 0  Charles Darwin's brother is one of the best roles

I've ever seen on a PBS documentary.

 Darwin married his cousin  finches can be blown really really far by the wind.  There were some really big rodents at one point

Thing I hoped you’d get from the video

• 1. Individuals evolve over generations.

Thing I hoped you’d get from the video

 2. Evolution is guided by fitness in a

given environment.

Thing I hoped you’d get from the video

• 3. Individuals are the “product” of their

parents.

Thing I hoped you’d get from the video

• 4. An individuals traits can undergo

random mutation

4

Thing I hoped you’d get from the video

• 5. Genotype vs. phenotype

Getting back to the video

• 6. Nature has no apparent

goal (except to create individuals suitable to a given environment).

Evolutionary Algorithms



An EA uses some mechanisms inspired by biological evolution: reproduction, mutation, recombination, natural selection and survival of the fittest.



Candidate solutions to the optimization problem play the role of individuals in a population, and the cost function determines the environment within which the solutions "live".



Evolution of the population then takes place after the repeated application of the above operators.

Evolution and Computing

 Your “world” is the context of a given

problem

Evolution

 You will be evolving generations of

“individuals”

 “Individuals” are solutions to your

problem

Survival of the fittest

 Evolution will be guided by fitness…  You must decide what makes a “fit”

solution

 Are you a good solution or a bad

solution.

5

Genotype / Phenotype

 Genotype (DNA) of your solution will be

a common data structure

 List, array, tree, etc.

 Phenotype will be an “individual”

solution derived from the DNA

 You must define this translation.

Solution breeding

 Child solutions are spawned from

parent solutions

 Crossover  Mechanism to combine DNA to create DNA

• f offspring

Solution mutation

 The DNA from a single individual can

randomly mutate.

 Mutation  Must define how given the data structure

representing the genotype.

Solution mutation

 The DNA from a single individual can

randomly mutate.

 Mutation  Must define how given the data structure

representing the genotype.

Goal

 To create individuals suitable to a given

environment…

 In other words…

 Create “fit” individuals  Create “good” solutions  Create solutions that best solve the

problem.

Questions so far

6

Evotionary Computation process

Initialize population Select individuals for crossover (based on fitness function Crossover Mutation Insert new offspring into population Are stopping criteria satisfied? Finish

Given a problem

 To use evolutionary algorithms your must:

 Define your genotype  Identify your phenotype  Define the genotype -> phenotype translation  Define crossover and mutation operators  Define fitness  Determine selection criteria  Set population parameters

 This course will explore each part of the

process in depth!

Lectures

 Lectures will cover:

 Process of using evolutionary algorithms  Process applied to my favorite problem

(TSP)

 Example uses of EAs in practice

 Listed as APPLICATION in schedule.  Guest lecturers.

Course deliverables

 In this course you will:

 Choose a hard, multidimensional problem

to solve.

 Define and code evolutionary methods to

solve the problem.

 Experiment with the parameters.

 You will be working in teams of 2.

Teaming up

 Find a common problem…  Mycourses interest questionaire:

problem/area you would like to solve.

 Please fill in by Monday.  I will list students and topics on Web by

next class.

Course deliverables

 Four major deliverables:

 Checkpoints / assignments  Implementation code  Final report  Presentation

 And for grad students:

 Grad report.

7

Checkpoints

 Weekly (approx) assignments, each exploring

• n task in using EC on your problem:

 1. Framework  2. Problem Statement  3. Genotype / Phenotype  4. Crossover / Mutation  5. Fitness / Selection  6. Population Parameters

Report

 Final findings of how well your EC

solved your problem.

 Summary of findings in checkpoint

Code

 Implementation of the EC:

 From scratch  Using a toolkit / API (some listed on Web

page)

 Use language of your choice  Readable code please

Presentation

 Oral summary of your final report  Last weeks of class + finals week.  Peer evaluation.

Grad report

 About grad work

 Help you with project / thesis  Report can be the background/previous

work section of your grad report

 Choose problem related to your work.

Grading

20% 15% Presentation 25% 20% Report 20% Grad Report 25% 20% Code 30% 25% Assignments Undergrad Graduate

8

Ground rules

 Use the language / toolkit of your

choice

 Toolkit use is optional

 Work in teams of 2.  All submissions will be done

electronically using mycourses dropboxes.

Due dates

 Checkpoints -- periodically (approx

weekly)

 Code + Report:  Presentations: week 9, 10, finals week  Grad report: finals week

Questions

 Next time:

 EA Frameworks  Defining your problem

 Please think about your problem and fill

• ut mycourses questionairre.