Outlier detection in high-dimensional data is a challenging yet importanttask, as it has applications in, e.g., fraud detection and quality control.State-of-the-art density-based algorithms perform well because they 1) take thelocal neighbourhoods of data points into account and 2) consider featuresubspaces. In highly complex and high-dimensional data, however, existingmethods are likely to overlook important outliers because they do notexplicitly take into account that the data is often a mixture distribution ofmultiple components. We therefore introduce GLOSS, an algorithm that performs local subspaceoutlier detection using global neighbourhoods. Experiments on synthetic datademonstrate that GLOSS more accurately detects local outliers in mixed datathan its competitors. Moreover, experiments on real-world data show that ourapproach identifies relevant outliers overlooked by existing methods,confirming that one should keep an eye on the global perspective even whendoing local outlier detection.
展开▼
