Lecture Notes

• Background:
  • David E. Goldberg's 1989 classic first textbook on Genetic Algorithms:  
    • Dave was my MS and Ph.D advisor at the University of Illinois in Urbana-Champaign.
    • Dave is director of the IlliGAL (Illinois Genetic Algorithms Laboratory), where I did my graduate work from 1992 to 1996.

Assignments

• Reading Assignment 1:  The Contreras Paper
  •  Our first CONFERENCE PAPER
  •  from the proceedings of GECCO 2007, held July, 2007 in London, UK
  •  (hardcopy handed out in class)

• HW1:   What's New in the Contreras Paper   (pass/fail)

• Reading Assignment 2:  
  • Our first JOURNAL PAPER
  • from the journal Artificial Life: 
  • This paper is a great read.  It extends the exciting work of Karl Sims, whose famous, groundbreaking videos we watched last week, and it includes links to new videos.
  • I want you to follow the link below, which goes to Chris Adami's web pages on his research with the Avida alife system, and then find and follow the links to the recent paper "Evolving Virtual Creatures and Catapults".   I want to see if you can download this paper for free.  (The NMU Lydia M. Olson Library has subscriptions to many on-line journals, and I want to see if this one works for you.  In general, journal subscriptions are expensive.   Usually, one has to use an inter-library loan service, specifically requesting a particular article, and then wait up to four weeks for a photocopy...   Obviously it is much better to have full-text, full-color,on-line access to an entire journal!)  
  • Download, if you can, and start perusing it for next Tuesday (Sept. 11) when we will start to discuss it.  You don't have to print it (up to you; I will print hardcopies for you if you ask me by email!).   You don't have to read the whole thing for Tues.  Remember, a good strategy, as we discussed in class, is to read the abstract, intro, conclusions, and/or discussion sections first (in any order you like!), look at images and tables next, then perhaps read "results", before finally reading the whole thing in the original, linear, order.  At least, this kind of "layering" works for me.  I might then quit after the first few "layers" if I determine the article is not pertinent to my current project.   Also, this is the most expeditious way to get some overall understanding of the paper in a short amount of time...
  • http://dllab.caltech.edu/pubs/