Detecting Unusual Input-Output Associations in Multivariate Conditional Data
This work addresses the need for outlier detection methods that can handle conditional data, which is important for applications where context-specific anomalies matter, but it is incremental as it builds on existing conditional outlier detection concepts.
The paper tackles the problem of detecting multivariate conditional outliers, which are unusual input-output associations in data, by introducing a novel framework that uses a decomposable conditional probabilistic model with weighted components based on reliability. Experimental results across various domains show the framework successfully identifies these outliers.
Despite tremendous progress in outlier detection research in recent years, the majority of existing methods are designed only to detect unconditional outliers that correspond to unusual data patterns expressed in the joint space of all data attributes. Such methods are not applicable when we seek to detect conditional outliers that reflect unusual responses associated with a given context or condition. This work focuses on multivariate conditional outlier detection, a special type of the conditional outlier detection problem, where data instances consist of multi-dimensional input (context) and output (responses) pairs. We present a novel outlier detection framework that identifies abnormal input-output associations in data with the help of a decomposable conditional probabilistic model that is learned from all data instances. Since components of this model can vary in their quality, we combine them with the help of weights reflecting their reliability in assessment of outliers. We study two ways of calculating the component weights: global that relies on all data, and local that relies only on instances similar to the target instance. Experimental results on data from various domains demonstrate the ability of our framework to successfully identify multivariate conditional outliers.