# Help needed on method to use for anomaly detection in Computer Science

I think people here could guide me in solving a problem related to anomaly detection. The term anomaly here refers to some undesired event occurring in the system like a virus infection.

I could get to know about it from more than one source. For example extracting value from two different data structure and if the value is different it is certain that virus infection is there.

In order to remove the false positive cases, information is gathered from different data structures or mechanisms. In that certain information are less trusted and certain information are more trusted.

I am looking for a mathematical method, that could easily handle this type of situations. Whether fuzzy/Genetic Algorithm/Neural Net fits here ? Found in some places they using one normality based approach(using z score). Please help.

-
Sounds like a typical machine learning problem. –  Raphael Jan 9 '11 at 14:10
This sounds more like a statistics / CS question. Have you tried asking on stats.stackexchange.com or cstheory.stackexchange.com? –  Qiaochu Yuan Jan 10 '11 at 16:48
I thought of asking here first. –  user5507 Jan 10 '11 at 16:49

You might want to take a look to one-class SVM method.

-
can you say a bit more about the method, or perhaps provide a link to an internet resource, or a citation to a textbook? –  Willie Wong Jan 26 '11 at 15:31
Here are some references: * B. Schölkopf, J. Platt, J. Shawe-Taylor, A. J. Smola, and R. C. Williamson. Estimating the support of a high-dimensional distribution. * Chapter 8 in the book "Learning with Kernels" by B. Schölkopf and A. J. Smola. –  TheBug Jan 27 '11 at 12:32