[jgap-users] example for custom chromosomes
Brought to you by:
klausikm
Menu
▾
▴
[jgap-users] example for custom chromosomes
|
From: Klaus M. <me...@br...> - 2006-06-08 19:40:39
|
Hello Susan, please find attached a possible implementation of your requirement (40 Genes with 4 fields, 1 Gene with 3 fields). The files must be but under a directory examples/distinctGenes (see their package declaration). Normally, a gene has no fields. To still enable fields within genes, I used a CompositeGene which in turn contains primitive genes. Thus, each primitive gene represents a "field". So the 41 "genes" are 40 CompositeGene's and the 4 resp. 4 "fields" are 4 resp 3 primitive genes within the CompositeGene's. I checked the whole stuff in to CVS. It works with JGAP 3.0 (accessible via CVS). If you use JGAP 2.6, you have to adapt the code slightly. If you don't come along with that adaptation, please let me know. But I recomment you check out the 3.0 version from CVS, if not already happened. Don't hesitate asking me if something is not as you require it. Best Klaus The German JUnit book for successfully testing software with JUnit: http://www.junit-buch.de _____ From: susan al-naqshbandi [mailto:sus...@ya...] Sent: Wednesday, June 07, 2006 6:05 PM To: Klaus Meffert Subject: RE: [jgap-users] chromosome representation The attachment is the paper I promiss you..... Klaus Meffert <me...@br...> wrote: Very interesting issue! Sure I am interested in the paper you mention :-) Best Klaus _____ From: susan al-naqshbandi [mailto:sus...@ya...] Sent: Tuesday, June 06, 2006 11:01 PM To: Klaus Meffert Subject: RE: [jgap-users] chromosome representation Hello Klaus Thank you for your reply....i really want to take the advantages of jgap to build my project ...I want to build an immune system for network Intrusion detection so one of the methods is by using genetic with fuzzy .....that what I want to depend on and build and if you like I will send you a paper that explain what i want to start with then i improve it in the way I need..... Best Regards Susan Klaus Meffert <me...@br...> wrote: Hello Susan, thanx for your interest in JGAP! Please give me one or two additional days to answer your good questions. Maybe could could try the CompositeGene until then. For your second question I think it would be a good solution using the fitness function to cope with static values. Presetting some values only makes overall sense, IMO, if there is no randomization of those distinguished values, i.e. they are static. And if they are, then the fitness function could assume them. May I ask why your problem requires this? Perhaps I could then give a more specific answer. Best Klaus _____ From: susan al-naqshbandi [mailto:sus...@ya...] Sent: Tuesday, June 06, 2006 12:40 AM To: jga...@li... Subject: [jgap-users] chromosome representation Hello, I want to ask if I can build a chromosome with 41 gene and each gene contain 4 field but not gene number 41 which should contains 3 field only......and can i input value of some fields of gene (i.e not all fields generate randomly)?and how can i represent it ..... Thank you and best regards Susan __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com |