| Presentation | 2021-03-19 [Encouragement Talk] Utility of Training Data in Sequential Accumulation Learning-Based Anomaly Detection Natsuki Fukazawa, Naoki Yoshida, Shingo Ata, Ikuo Oka, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In Network-based Intrusion Detection Systems (NIDS) using supervised learning, one of important challengings is how to collect and accumlate good learning data to achieve high detection accuracy. So far we proposed a mechanism to accumlate learning data concecutively by associating events of em honeypots and flow characteristics of monitored packets. It is expected that our system can detect anomalies more accurate by running the system longer because the volume of learning data becomes larger. However, it is still unclear the policy how to accumlate the learning data efficiently in terms of the accuracy of detection. In this paper, we conduct quantitative evaluation how an accumulate policy has an impact to the accuracy of detection. We investigate the relation of accumlated learning data and the performance of anomaly detection. Through this paper we aim to consider a guideline to the efficient way of learning data accumulation. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Anomaly Detection / Traffic Pattern / Honeypot / Machine Learning / Attack Classification |
| Paper # | ICM2020-69 |
| Date of Issue | 2021-03-11 (ICM) |
| Conference Information | |
| Committee | ICM |
|---|---|
| Conference Date | 2021/3/18(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Online |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Kazuhiko Kinoshita(Tokushima Univ.) |
| Vice Chair | Yoichi Sato(OSL) / Haruo Ooishi(NTT) |
| Secretary | Yoichi Sato(NTT) / Haruo Ooishi(Bosco) |
| Assistant | Tetsuya Uchiumi(Fujitsu Lab.) |
| Paper Information | |
| Registration To | Technical Committee on Information and Communication Management |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | [Encouragement Talk] Utility of Training Data in Sequential Accumulation Learning-Based Anomaly Detection |
| Sub Title (in English) | |
| Keyword(1) | Anomaly Detection |
| Keyword(2) | Traffic Pattern |
| Keyword(3) | Honeypot |
| Keyword(4) | Machine Learning |
| Keyword(5) | Attack Classification |
| 1st Author's Name | Natsuki Fukazawa |
| 1st Author's Affiliation | Osaka City University(Osaka City Univ.) |
| 2nd Author's Name | Naoki Yoshida |
| 2nd Author's Affiliation | Osaka City University(Osaka City Univ.) |
| 3rd Author's Name | Shingo Ata |
| 3rd Author's Affiliation | Osaka City University(Osaka City Univ.) |
| 4th Author's Name | Ikuo Oka |
| 4th Author's Affiliation | Osaka City University(Osaka City Univ.) |
| Date | 2021-03-19 |
| Paper # | ICM2020-69 |
| Volume (vol) | vol.120 |
| Number (no) | ICM-433 |
| Page | pp.pp.52-57(ICM), |
| #Pages | 6 |
| Date of Issue | 2021-03-11 (ICM) |