Isolation-based Spherical Ensemble Representations for Anomaly Detection

Anomaly detection is a critical task in data mining and management with applications spanning fraud detection, network security, and log monitoring. Despite extensive research, existing unsupervised anomaly detection methods still face fundamental challenges including conflicting distributional assumptions, computational inefficiency, and difficulty handling different anomaly types. To address these problems, we propose ISER (Isolation-based Spherical Ensemble Representations) that extends existing isolation-based methods by using hypersphere radii as proxies for local density characteristics while maintaining linear time and constant space complexity. ISER constructs ensemble representations where hy-persphere radii encode density information: smaller radii indicate dense regions while larger radii correspond to sparse areas. We introduce a novel similarity-based scoring method that measures pattern consistency by comparing ensemble representations against a theoretical anomaly reference pattern. Additionally, we enhance the performance of Isolation Forest by using ISER and adapting the scoring function to address axis-parallel bias and local anomaly detection limitations. Comprehensive experiments on 22 real-world datasets demonstrate ISER's superior performance over 11 baseline methods. Anomaly detection is the task of identifying data points that deviate significantly from the majority of observations, with applications in fraud detection, network security, and quality control (Chandola et al., 2009; Liu et al., 2024; Tang et al., 2024; Song et al., 2023). Despite extensive research, developing effective unsupervised anomaly detection methods remains challenging due to several fundamental limitations. Existing methods face a critical trade-off between computational efficiency and handling varying local densities. Density-based methods like Local Outlier Factor (Breunig et al., 2000) address this but require quadratic time complexity, limiting scalability.