| Presentation | 2017-06-16 Inferring causal parameters of anomalies detected by autoencoder using sparse optimization Yasuhiro Ikeda, Keisuke Ishibashi, Yusuke Nakano, Keishiro Watanabe, Ryoichi Kawahara, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | The anomaly detection algorithm based on an autoencoder has attracted much attention. An autoencoder is a neural network model used for unsupervised learningand requires only data in normal time as training data to output abnormality of test dataaccording to how far they are different from the training data. The autoencoder therefore seems to be desirable as an anomaly detection algorithmunder the situation that abnormal data cannot be obtained sufficiently. However, identifying the root cause of the anomalies detected by the autoencoder is difficultsince the causal input parameters of the anomalies are not directly indicated. In this paper, we propose an algorithm for inferring causal input parameters of an autoencoder for anomaliesby using sparse optimization. We also evaluate the algorithm through simulated data and network benchmark data. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | deep Learning / autoencoder / cause estimation |
| Paper # | IN2017-18 |
| Date of Issue | 2017-06-08 (IN) |
| Conference Information | |
| Committee | IN |
|---|---|
| Conference Date | 2017/6/15(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Roudou-Fukushi-Kaikan (Koriyama) |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Katsunori Yamaoka(Tokyo Inst. of Tech.) |
| Vice Chair | Takuji Kishida(NTT) |
| Secretary | Takuji Kishida(NTT) |
| Assistant | Hiroaki Karasawa(NTT) / 植田 一暁(KDDI R&D Labs.KDDI R&D Labs.) |
| Paper Information | |
| Registration To | Technical Committee on Information Networks |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Inferring causal parameters of anomalies detected by autoencoder using sparse optimization |
| Sub Title (in English) | |
| Keyword(1) | deep Learning |
| Keyword(2) | autoencoder |
| Keyword(3) | cause estimation |
| 1st Author's Name | Yasuhiro Ikeda |
| 1st Author's Affiliation | Nippon Telegraph and Telephone Corporation(NTT) |
| 2nd Author's Name | Keisuke Ishibashi |
| 2nd Author's Affiliation | Nippon Telegraph and Telephone Corporation(NTT) |
| 3rd Author's Name | Yusuke Nakano |
| 3rd Author's Affiliation | Nippon Telegraph and Telephone Corporation(NTT) |
| 4th Author's Name | Keishiro Watanabe |
| 4th Author's Affiliation | Nippon Telegraph and Telephone Corporation(NTT) |
| 5th Author's Name | Ryoichi Kawahara |
| 5th Author's Affiliation | Nippon Telegraph and Telephone Corporation(NTT) |
| Date | 2017-06-16 |
| Paper # | IN2017-18 |
| Volume (vol) | vol.117 |
| Number (no) | IN-89 |
| Page | pp.pp.61-66(IN), |
| #Pages | 6 |
| Date of Issue | 2017-06-08 (IN) |