A Survey on Artificial Immune System

by Akira Imada (February 2004)

This is a survey on Artificial Immune System I've briefly made via http://www.google.com. This is quite a new approach to the "Computational Intelligence". At the moment, on the 15 February 2004, when I explored the site with keyworda of "artificial immune system & pdf", 79 pages were there from which I ftp'ed 45 papers and I put them all here in our web page. You might just crick the title of the paper which is of your interest, if any. Please feel free to send e-mails to akira@bstu.by for your comments, questions, bug-reports, further-surevey, experiment-with-toy-example or whatever else. Those are very welcome in order to elaborate this document. Sorry for no-consistency. That is, some papers have shown the year of publication while others are not; some papers are given all the author's names while others are just with "et al." or something like that. This is because they are explored via google and I didn't get such an information sometimes. If the authors are from a sort of center of research in this field of AIS, I add the name of the institute/university, but not exhaustively. Anyway the page will be modified for the time being. Please enjoy.

What is Artificial Immune System?

Artificial Immune System (AIS) is a computational technique Our Immune system has a fancy capability to protect us from an attack by viruses. Employing its collective behavior, it memorizes an information of viruses that previously attacked us. We borrow the idea from nature, that is to say, AIS is one of biologically-inspired-computational-technologies.

Application to Steganography

Some of our group have already applied Neural Networks to Steganography. So, this would be a good paper to start with.

Detecting Virus Invasion to Network System.

When we listen to the term Artificial Immune System, this topic is most likely to reminiscent of for most of us. As a result we have a whale of big amount of papers on this topic.

Applications to Multi-Agent-System

It is also very natural to imagine a Multi Agent System when we thought of AIS.

Anomaly Detection.

Personally I'm currently interested in this "Anomaly Detection by AIS" thinking of applying this anomaly detection to my problem of "one already known peak in a huge almost everywhere completley flat fitness landscape.

Simulation of Human Memory.

The other of my special concern is a consideration if we can use AIS to model human memory, which used to be very popular topic in Artificial Neural Network community. Here I put only one article but if I have a time I will further survey on this topic.

Application to the other practical domains.

More Theoretical/Biological Approaches.

The bellow listed are more theoretical and/or more biological works