| Presentation | 2019-11-02 On Robustness of Machine-Learning-Based Malware Detection Wanjia Zheng, Kazumasa Omote, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | As the 2020 Tokyo Olympics are approaching, the possibility of being targeted by attackers has further increased in Japan. In order to prevent cyber attacks, machine learning technology that is spreading in various fields including images and voices recognition is also being used in the cyber security field. However, there are many adversarial attack methods that trick machine learning models and cause it to malfunction. In this study, we propose a robust malware detection model that is strong against adversarial attack methods combining dimensional compression and machine learning, and evaluates robustness using PE file's information from FFRI Dataset 2018. Measure the effectiveness and superiority of the proposed method from the viewpoint of robustness other than detection accuracy by measuring the distance(Euclidean distance between centroids, etc.) between malware and normal file data and evaluating the success rate of the attack method (SVM-Attack) is shown in this study. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Malware Detection / Machine Learning / Dimension Reduction / Adversarial Attack |
| Paper # | ISEC2019-83,SITE2019-77,LOIS2019-42 |
| Date of Issue | 2019-10-25 (ISEC, SITE, LOIS) |
| Conference Information | |
| Committee | ISEC / SITE / LOIS |
|---|---|
| Conference Date | 2019/11/1(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Osaka Univ. |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Shiho Moriai(NICT) / Tetsuya Morizumi(Kanagawa Univ.) / Tomohiro Yamada(NEL) |
| Vice Chair | Shoichi Hirose(Univ. of Fukui) / Tetsuya Izu(Fujitsu Labs.) / Masaru Ogawa(Kobe Gakuin Univ.) / Takushi Otani(Kibi International Univ.) / Toru Kobayashi(Nagasaki Univ.) |
| Secretary | Shoichi Hirose(NICT) / Tetsuya Izu(Tsukuba Univ.) / Masaru Ogawa(Toyo Eiwa Univ.) / Takushi Otani(KDDI Research) / Toru Kobayashi(Research Organization of Information and Systems) |
| Assistant | Dai Yamamoto(Fujitsu Labs.) / Yuuji Suga(IIJ) / Nobuyuki Yoshinaga(Yamaguchi Pref Univ.) / Daisuke Suzuki(Hokuriku Univ.) / Kenichi Arai(Nagasaki Univ.) |
| Paper Information | |
| Registration To | Technical Committee on Information Security / Technical Committee on Social Implications of Technology and Information Ethics / Technical Committee on Life Intelligence and Office Information Systems |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | On Robustness of Machine-Learning-Based Malware Detection |
| Sub Title (in English) | |
| Keyword(1) | Malware Detection |
| Keyword(2) | Machine Learning |
| Keyword(3) | Dimension Reduction |
| Keyword(4) | Adversarial Attack |
| 1st Author's Name | Wanjia Zheng |
| 1st Author's Affiliation | University of Tsukuba(U. Tsukuba) |
| 2nd Author's Name | Kazumasa Omote |
| 2nd Author's Affiliation | University of Tsukuba(U. Tsukuba/NICT) |
| Date | 2019-11-02 |
| Paper # | ISEC2019-83,SITE2019-77,LOIS2019-42 |
| Volume (vol) | vol.119 |
| Number (no) | ISEC-257,SITE-258,LOIS-259 |
| Page | pp.pp.133-140(ISEC), pp.133-140(SITE), pp.133-140(LOIS), |
| #Pages | 8 |
| Date of Issue | 2019-10-25 (ISEC, SITE, LOIS) |