An Apperceived Method of Danger Signals for Computer Systems Based on Cloud Model
He Yang, Yiwen Liang · 2009
Danger Theory is a novel method of Biological Immunology. Artificial Immune Systems researchers may extract benefits from the theory, especially in anomaly detection. The definition of danger signals is one of the most important problems in Danger Theory. For the distinction between danger and safety is fuzzy and precarious, the precise calculation method is not suit for this problem. Cloud Model is an effective tool to transform qualitative concepts into quantitative expressions for uncertain problems. In this paper, a suggestive definition of danger signals based on Cloud Model is presented. The changes of key features of a computer system are collected and integrate by the rules generator based on Cloud Model, and danger signals are presented. The experimental results demonstrated that the proposed definition of danger signals is feasible.