Showing 1–1 of 1 results for author: Lindskou, M

Detecting Outliers in High-dimensional Data with Mixed Variable Types using Conditional Gaussian Regression Models

Authors: Mads Lindskou, Torben Tvedebrink, Poul Svante Eriksen, Niels Morling

Abstract: Outlier detection has gained increasing interest in recent years, due to newly emerging technologies and the huge amount of high-dimensional data that are now available. Outlier detection can help practitioners to identify unwanted noise and/or locate interesting abnormal observations. To address this, we developed a novel method for outlier detection for use in, possibly high-dimensional, dataset… ▽ More Outlier detection has gained increasing interest in recent years, due to newly emerging technologies and the huge amount of high-dimensional data that are now available. Outlier detection can help practitioners to identify unwanted noise and/or locate interesting abnormal observations. To address this, we developed a novel method for outlier detection for use in, possibly high-dimensional, datasets with both discrete and continuous variables. We exploit the family of decomposable graphical models in order to model the relationship between the variables and use this to form an exact likelihood ratio test for an observation that is considered an outlier. We show that our method outperforms the state-of-the-art Isolation Forest algorithm on a real data example. △ Less

Submitted 19 May, 2021; v1 submitted 3 March, 2021; originally announced March 2021.