236 Chapter 8 GenetiC alGorithms
The advantages and disadvantages of genetic algorithms are similar
to those of artificial neural nets. The major disadvantages are the chance-
dependent outcome and lengthy computation time.
8.13 WARM-UP QUESTIONS, EXERCISES, AND PROJECTS
Warm-up Questions
What is evolutionary computing?1.
What is evolutionary strategy?2.
Illustrate the general computing flow in a genetic algorithm. 3.
Why do genetic algorithms work?4.
Describe the various steps in designing a genetic programming solu-5.
tion.
What two conditions should a problem satisfy in order to solve it by a 6.
genetic algorithm?
Consider the problem of finding the shortest path through different 7.
towns, such that each town is visited only once and in the end return to
the starting town. Let us assume that we use a genetic algorithm to solve
this problem, in which genes represent links between pairs of towns.
How many genes will be used in a chromosome in each individual if the
number of towns is 20?
Visit the website http://www.obitko.com/tutorials/genetic-algorithms/8.
tsp-example.php to see a genetic algorithm applet on the famous Trav-
eling Salesperson Problem. Attempt to run the genetic algorithm with
different crossovers and mutations. Study how the genetic algorithm
performs.
Limitations with typical genetic algorithms have led researchers to look 9.
for a more powerful evolutionary system. Keeping this in mind, identify
four weaknesses of genetic algorithms.
Use a freeware tool from the Microsoft research website at http:// 10.
research.microsoft.com/en-us/um/redmond/groups/adapt/msbnx/
and experiment with Bayesian networks.
76473_CH08_Akerkar.indd 236 8/11/09 10:19:09 AM
8.13 Warm-up Questions, exerCises, and projeCts 237
Exercises
Construct a genetic algorithm to train a feed-forward neural network.1.
Explore the use of a genetic algorithm to cluster data.2.
Develop sample programs from the genetic cycle shown in Figure 8.2.3.
How can genetic algorithms be employed to optimize the design of 4.
neural networks?
How are genetic algorithms used when searching for a solution to a 5.
problem? Consider the following problem: you are deciding what CDs
to buy. There are six different ones that you are considering. Some are
cheaper than others, and some are your favorites. How can a possible
solution be represented as a bit string (sequence of zeros and ones)?
In Exercise 5, how could genetic algorithms be used to determine pos-6.
sible solutions?
Given the following data on insurance risk, explain how a suitable 7.
classification rule might be found using neural networks and genetic
algorithms:
Example City Age Gender Risk
1 New York Young M Low
2 New York Young F Low
3 Boston Old M Low
4 Boston Young M High
5 Boston Young F High
Projects
Search for material on List Processing (LISP), a high-level functional 1.
language, and write a report on how LISP is most suitable for genetic
programming.
Design a genetic algorithm to build a simple neural network.2.
76473_CH08_Akerkar.indd 237 8/11/09 10:19:09 AM
238 Chapter 8 GenetiC alGorithms
References
Bonsma, E., Shackleton, M., & Shipman, R. Eos—An Evolutionary and Ecosys-
tem Research Platform, 2000, BT Technology Journal (2000) 18, pp. 24–31.
Hales, D. Introduction to Genetic Algorithms, 2006. Retrieved from http://cfpm.
org/~david/talks/ga2006/ga-pres6.ppt
Hayes, G. Genetic algorithm and genetic programming, 2007. Retrieved from http://
www.inf.ed.ac.uk/teaching/courses/gagp/slides07/gagplect6.pdf
Holland, J. H. Adaptation in natural and artificial systems, Ann Arbor, Michigan:
The University of Michigan Press; 1975.
Jong, K. D. Evolutionary Computation: A Unified Approach, Cambridge, MA: MIT
Press; 2006.
http://www.solver.com/
Mitchell, M. An Introduction to Genetic Algorithms. Cambridge, MA: MIT Press;
1996.
Obitko, M. Introduction to Genetic Algorithms, 1998. Retrieved from http://www.
obitko.com/tutorials/genetic-algorithms/
Potter, M. The Design and Analysis of a Computational Model of Cooperative Coevo-
lution, PhD Thesis, Fairfax, VA: George Mason University; 1997.
Sipper, M. A brief introduction to genetic algorithms, 1996. Retrieved from http://
www.cs.bgu.ac.il/~sipper/
Weck, O. D. Heuristic techniques: A basic introduction to genetic algorithms, 2004. Re-
trieved from http://ocw.mit.edu/NR/rdonlyres/Aeronautics-and-Astronautics/
16-888Spring-2004/D66C4396-90C8-49BE-BF4A-4EBE39CEAE6F/0/
MSDO_L11_GA.pdf
76473_CH08_Akerkar.indd 238 8/11/09 10:19:09 AM

Get Knowledge-Based Systems now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.