LSCP: Locally Selective Combination in Parallel Outlier Ensembles

Abstract

In unsupervised outlier ensembles, the absence of ground truth makes the combination of base outlier detectors a challenging task. Specifically, existing parallel outlier ensembles lack a reliable way of selecting competent base detectors, affecting accuracy and stability, during model combination. In this paper, we propose a framework---called Locally Selective Combination in Parallel Outlier Ensembles (LSCP)---which addresses the issue by defining a local region around a test instance using the consensus of its nearest neighbors in randomly selected feature subspaces. The top-performing base detectors in this local region are selected and combined as the model's final output. Four variants of the LSCP framework are compared with seven widely used parallel frameworks. Experimental results demonstrate that one of these variants, LSCP_AOM, consistently outperforms baselines on the majority of twenty real-world datasets.

arXiv commentProceedings of the 2019 SIAM International Conference on Data Mining (SDM)

References in corpus (3)

• Learning Representations for Outlier Detection on a Budget
• Less is More: Building Selective Anomaly Ensembles
• DCSO: Dynamic Combination of Detector Scores for Outlier Ensembles

Cited by in corpus (16)

• ECOD: Unsupervised Outlier Detection Using Empirical Cumulative Distribution Functions
• COPOD: Copula-Based Outlier Detection
• Negative Selection Algorithm Research and Applications in the last decade: A Review
• SUOD: Accelerating Large-Scale Unsupervised Heterogeneous Outlier Detection
• Fuzzy Granule Density-Based Outlier Detection with Multi-Scale Granular Balls
• PyOD: A Python Toolbox for Scalable Outlier Detection
• fSEAD: a Composable FPGA-based Streaming Ensemble Anomaly Detection Library
• Combining Machine Learning Models using combo Library
• Music Artist Classification with Convolutional Recurrent Neural Networks
• Comparison of Outlier Detection Techniques for Structured Data
• SynC: A Unified Framework for Generating Synthetic Population with Gaussian Copula
• Ensemble Grammar Induction For Detecting Anomalies in Time Series
• Wavelet-based temporal models of human activity for anomaly detection in smart robot-assisted environments
• Modeling Heterogeneous Statistical Patterns in High-dimensional Data by Adversarial Distributions: An Unsupervised Generative Framework
• IPOF: An Extremely and Excitingly Simple Outlier Detection Booster via Infinite Propagation
• BikNN: Anomaly Estimation in Bilateral Domains with k-Nearest Neighbors