A Computer Virus Detection System Based on Smooth Support Vector Machine

• Main Authors: Chan-Wei Chen, 陳展維
• Other Authors: Shi-Jinn Horng
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/45566772032487689328

碩士 === 國立臺灣科技大學 === 資訊工程系 === 92 === In this paper we purpose a computer virus detection system based on smooth support vector machine(SSVM). This method is not like the traditions method that use virus code to detect virus. We pick the virus features and use smooth support vector machine to training then we can get a good classifier. Through the experiment prove our system has 100% detection rate for the known virus and 95% detection rate for the unknown virus. Our research is focus on Microsoft Windows OS. Every Executable file on Microsoft Windows OS has Portable Executable Table(PE Table). It include the dll files,API Function and API parameter information. In the paper, we first use the features in PE Table and use SSVM to training those features. Through the fast machine learning ability of the SSVM to design a computer virus detection system.