Location:Home > Engineering science > Computer Science > Computer System Architecture > Research of Virus Detection Methods Based on Multiple Anti-virus Softwares Collaboration
Details
Name

Research of Virus Detection Methods Based on Multiple Anti-virus Softwares Collaboration

Downloads: []
Author
Tutor: GuGuoChang
School: Harbin Engineering University
Course: Computer System Architecture
Keywords: multiple anti-virus softwares collaboration,virus detection,BP neural network
CLC: TP309.5
Type: Master's thesis
Year:  2011
Facebook Google+ Email Gmail Evernote LinkedIn Twitter Addthis

not access Image Error Other errors

Abstract:
The rapid development of Internet accelerate the computer viruse dramatically, but the virus analyst for the identification and investigation of computer viruses is still manual. Face a lot of suspicious files to be processed and to make a rapid response, it is very necessary to process a large number of files on the server. During the pretreatment, detection rate and false positives are the most important target. Although the existing anti-virus software attach the two target important and trying to do their best, the two target is always conflict.Neural network has got widespread attention for its specialties, such as parallel processing, adaptive organization, associative memory and robustness. BP neural network is an important model of the artificial neural network. It have extensive use in the aspects such as character recognition, pattern classification, conversion from text to voice, image compression and decision support.On the basis of in-depth understanding the process of the virus file detection and the main technology adopted, first of all, it put forward in this article the method of the pre-process for the large number of the files to be pre-thought by means of multiple anti-virus software work together. This method can increase the rate of virus detection, as well as control the false alarm in suitable range. Secondly, in the process of the anti-virus softwares work together, with the combination of BP neural network will make the final decision more accurate. Meanwhile, for the need of the virus detection, make improvement for the shortcomings of the easy forgetting of the old samples and the easy falling into the local minimum, which makes the proposed method of virus detection more reasonable. Finally, across the experiment for the virus has been verified on the file and the file produced false detection, validate the method proposed.
Related Dissertations
Last updated
Sponsored Links
Home |About Us| Contact Us| Feedback| Privacy | copyright | Back to top