Evolutionary Algorithms
“The genes are the master programmers, and they are programming for their lives.” –Richard Dawkins
 

 

Note: Final grades are available.

A few concluding remarks:

The submitted projects were of a very high quality, as many of you actually did a mini-thesis on the subject! I very much enjoyed teaching this course, and I m happy to learn from your projects that you also enjoyed the subject.

During this course, we saw a wide range of problems for which evolutionary algorithms provide efficient solutions. It is said that when the only tool you have is a hammer, every problem looks like a nail. During the course we saw that for many problems the key part is converting the problem into a chromosome and finding a suitable fitness function. In other words, having a hammer called EA, we actually tried to convert every problem into a nail (and mostly succeeded!). Of course EA has its shortcomings as well, and cannot solve every problem. But it is usually worth the effort.

My goal in this course was to add another tool called EA to your problem-solving toolkit. If next time you ponder on a problem, it passes your mind that you can try EA on it, then I have achieved my goal.

Thanks again to all 70+ of you for attending the course, and good luck!

--------------------------------------------

I. Welcome

Finally the course's webpage is up and running. On this page you can find the course syllabus and projects, and also some links to interesting EA-related content.

II. Syllabus

Here is the list of topics we will cover in the course: syllabus.

There will be two programming projects (50% of the grade) and the final exam will comprise the remaining 50%.

You can find Ant Colony Optimization presentation given by Israel Beniaminy here.

III. Forum

Someone was kind enough to open this forum on zoombut website, so here is the unofficial course forum:

http://www.zoombut.com/modules.php?name=Forums&showforum=1024

I will try to post there whenever I can, especially if there are interesting questions there (by "interesting" I don't mean "how do I compile this code?" sort of questions... that's what Google is for!).

IV. Projects

There will be two programming projects. You can submit them in pairs.

  • Project 1 (due: 14.12.08, extended until 16.12.08)
     
  • Project 2 (due: 1.2.09 extended until 29.3.09)

And one more thing, these are NOT "programming exercises" where you just write a bunch of code, compile it, and halleluiah it is working... The idea behind each part is that you experiment with it, change the parameters, play with all sort of bells and whistles and see how they affect the algorithms... Take my word for it, it is quite fun when you get the hang of it!

V. Grades

Final grades are available at the following links. If your grade does not appear, please let me know as soon as possible.

Project grades

Exam grades

VI. Links

A fascinating talk by Jeff Hawkins (who has written a very interesting book titled "On Intelligence"). I highly recommend that you check the TED website for some other great talks.
 

Ynet: אבולוציה בשחור-לבן: אלגוריתמים גנטיים בשירות השחמט

Galileo, no. 119, July 2008

Code and Demo Projects:

GA Articles from New Scientist:

 

 
 

© 2009 Omid David