admin管理员组文章数量:1609966
2013.ECML PKDD.Local Outlier Detection with Interpretation
- paper
- main idea
- contribution
- method
- Neighboring Set Selection
- Anomaly Degree Computation
- Outlier Interpretation
- citation
- experiment
paper
main idea
In LODI, we develop an approach that explores the quadratic entropy to adaptively select a set of neighboring instances, and a learning method to seek an optimal subspace in which an outlier is maximally separated from its neighbors.
We show that this learning task can be solved via the matrix eigen-decomposition and its solution contains essential information to reveal features that are most important to interpret the exceptional properties of outliers.
contribution
We propose a technique relying on the information theoretic measure of entropy to select an appropriate set of neighboring objects of an outlier candidate.
We then develop a method, whose solution firmly relies on the matrix eigen-decomposition, to learn an optimal one-dimensional subspace in which an outlier is most distinguishable from its neighboring set.
method
Neighboring Set Selection
Anomaly Degree Computation
Outlier Interpretation
citation
Local Outlier Detection with Interpretation (LODI) [37] and Local Outliers with Graph Projection (LOGP) [38] identify outliers in subspaces of the original feature space and in subspaces of a transformation of the original feature space respectively, like SOD and COP that were introduced in the previous section. But LODI and LOGP provide weights quantifying the importance of each identified feature.
experiment
本文标签: PKDDECMLLocalinterpretationDetection
版权声明:本文标题:2013.ECML PKDD.Local Outlier Detection with Interpretation 内容由热心网友自发贡献,该文观点仅代表作者本人, 转载请联系作者并注明出处:https://www.elefans.com/dongtai/1728575925a1164468.html, 本站仅提供信息存储空间服务,不拥有所有权,不承担相关法律责任。如发现本站有涉嫌抄袭侵权/违法违规的内容,一经查实,本站将立刻删除。
发表评论