CIVILICA We Respect the Science
(ناشر تخصصی کنفرانسهای کشور / شماره مجوز انتشارات از وزارت فرهنگ و ارشاد اسلامی: ۸۹۷۱)

Malware detection by converting to image using transfer learning in deep network

عنوان مقاله: Malware detection by converting to image using transfer learning in deep network
شناسه ملی مقاله: ITCT09_035
منتشر شده در نهمین کنفرانس بین المللی فناوری اطلاعات،کامپیوتر و مخابرات در سال 1399
مشخصات نویسندگان مقاله:

Reza Momen - Department of Computer Science and Engineering Apadana Institute Of Higher Education- Apadana Institute Of Higher Education
Seyed Mehdi Hazrati Fard - Department of Computer Science and Engineering Shiraz University

خلاصه مقاله:
For many years, various malware has been produced for malicious operations on computer systems. In recent years, it has become difficult to identify malware, because with the advancement of algorithms and software, malware detection and scanning of malware writers and developers have grown. Various and intelligent devices have been used to produce their malware, which makes it very difficult to identify malware. Todays, machine learning techniques are used to identify malware. In this article, using the technique of converting files to images and using neural network is the identification of malware. The main purpose of this article is to improve the identification of malware by transfer learning from one group of malware to another. We have been able to increase the accuracy from 83% to 95%.

کلمات کلیدی:
Malware , Machine learning , Neural network , Transfer learning (TL)

صفحه اختصاصی مقاله و دریافت فایل کامل: https://civilica.com/doc/1041353/