Detecting outliers in high dimensional datasets with examples (データベースシステム)
スポンサーリンク
概要
- 論文の詳細を見る
Many real applications of outlier detection can be utilized in various fields. Most of such applications process high dimensional datasets. In one of our previous work, we proposed an Example-Based outlier detection method making use of outlier examples to detect outliers that have similar "outlier-ness" to these examples in high dimensional datasets. However, it employs meshes for deciding outliers, and all points within the same cell are regarded as normal objects or outliers. Thus, sometimes normal objects may be detected as outliers, and vice versa. Distance-Based outlier (DB-Outlier) detection is a simple and commonly used approach, besides it can detect real outliers by calculating the distance between data points. In this paper, we introduce a method to detect DB-Outliers in high dimensional datasets based on outlier examples, and show some comparison experiments between our new method and the old one.
- 一般社団法人情報処理学会の論文
- 2007-07-02
著者
-
KITAGAWA Hiroyuki
Graduate School of Environmental Studies, Nagoya University
-
Kitagawa Hiroyuki
Graduate School Of Environmental Studies Nagoya University
-
Kitagawa Hiroyuki
Graduate School Of Systems And Information Engineering:center For Computational Sciences University
-
LI Yuan
Graduate School of Systems and Information Engineering, University of Tsukuba
-
Kitagawa Hiroyuki
Graduate School Of Systems And Information Engineering:center For Computational Sciences University
-
Li Yuan
Graduate School Of Systems And Information Engineering University Of Tsukuba
関連論文
- Social Bookmarking Induced Active Page Ranking
- Environmental magnetic record and paleosecular variation data for the last 40kyrs from the Lake Biwa sediments, Central Japan
- Atmospheric radiocarbon calibration curve beyond 12.4 cal kyr BP(Proceedings of the 19^ Symposium on Chronological Studies at the Nagoya University Center for Chronological Research in 2006,Part1)
- An algorithm for parallel holistic twig joins on a PC cluster (データベースシステム)
- Detecting outliers in high dimensional datasets with examples (データベースシステム)
- 2T-6 A Robust Method of Detecting DB-Outliers in High Dimensional Datasets
- 3R-9 Keyword Search Including Metadata in Relational Databases
- Detecting outliers in high dimensional datasets with examples (データ光学)
- 3T-2 Continuous Query over Uncertain Data Streams
- Social Bookmarking Induced Active Page Ranking
- MV-OPES : Multivalued-Order Preserving Encryption Scheme : A Novel Scheme for Encrypting Integer Value to Many Different Values
- Cube-Based Analysis for Maintaining XML Data Partition for Holistic Twig Joins
- Example-Based Outlier Detection for High Dimensional Datasets
- Interactive Outlier Detection Adaptive to Users' Intentions (夏のデータベースワークショップDBWS2004)
- Interactive Outlier Detection Adaptive to Users' Intentions (夏のデータベースワークショップ(DBWS2004))
- Example-Based Outlier Detection for High Dimensional Datasets
- Querying Topic Evolution in Time Series Document Clusters
- Example-Based Outlier Detection for High Dimensional Datasets