GUC: Unsupervised non-parametric Global Clustering and Anomaly Detection

Anomaly Detection is a crucial task in the fields of optimization and Machine Learning,with the ability of detecting global anomalies being of particular importance. In this paper,we propose a novel non-parametric algorithm for automatically detecting global anomaliesin an unsupervised manner. Our algorithm is both effective and efficient, requiring no priorassumptions or domain knowledge to be applied. It features two modes that utilize thedistance from the dataset’s center for grouping data points together. The first mode splitsthe dataset into global clusters where each cluster signifies proximity from the center. Thesecond mode employs a threshold value for splitting the points into outliers and inliers. Weevaluate our proposal against other prominent methods using synthetic and real datasets.Our experiments demonstrate that the proposed algorithm achieves state-of-the-art performance with minimum computational cost, and can successfully be applied to a wide rangeof Machine Learning applications.