Name
Abstract | ||
|---|---|---|
In wireless sensor networks (WSNs), there are many challenges for outlier detection, such as fault detection, fraud detection, intrusion detection, and so on. In this paper, the participation degree of instances in the hierarchical clustering process infers the relationship between instances. However, most of the existing algorithms ignore such information. Thus, we propose a novel fault detection technique based on the participation degree, called fault detection based on participation degree (FDP). Our algorithm has the following advantages. First, it does not need data training in labeled datasets; in fact, it uses the participation degree to measure the differences between fault points and normal points without setting distance or density parameters. Second, FDP can detect global outliers without local cluster influence. Experimental results demonstrate the performance of our approach by applying it to synthetic and real-world datasets and contrasting it with four well-known techniques: isolation forest (IF), local outlier factor (LOF), one-class support vector machine (OCS), and robust covariance (RC). |
Year | DOI | Venue |
|---|---|---|
2019 | 10.3390/s19071522 | SENSORS |
Keywords | Field | DocType |
outlier detection,fault detection,participation degree,hierarchical clustering,WSNs | Hierarchical clustering,Data mining,Local outlier factor,Anomaly detection,Fault detection and isolation,Support vector machine,Outlier,Electronic engineering,Engineering,Intrusion detection system,Wireless sensor network | Journal |
Volume | Issue | ISSN |
19 | 7.0 | 1424-8220 |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
6 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Zhang Wei | 1 | 392 | 53.03 |
| Gongxuan Zhang | 2 | 94 | 19.89 |
| Xiaohui Chen | 3 | 1 | 0.69 |
| Xiumin Zhou | 4 | 1 | 0.69 |
| Yueqi Liu | 5 | 1 | 0.69 |
| Junlong Zhou | 6 | 7 | 3.13 |