Improving Isolation Forest with Relative Mass
Kai Ming Ting
Doctor
Federation University Australia
Abstract: Isolation Forest, first introduced in 2008．is an anomaly detector with three unique features. First, it isolates anomalies, instead of profiling the norm which is the commonly used approach. Second, it does not use a distance measure. Third, it requires only a small training sample to produce a high-performing detector. As a result, Isolation Forest is one of the fastest anomaly detectors and one of the few that can easily scale up to big data. This talk provides a brief introduction to Isolation Forest and its existing weaknesses in anomaly detection and information retrieval. We show that these weaknesses can be overcome using relative mass. Isolation technique's relation to mass estimation will also be provided.
Bio: After receiving his PhD from the University of Sydney,Kai Ming Ting had worked at the University of Waikato, Deakin University and Monash University. He joins Federation University Australia since 2014. He had previously held visiting positions at Osaka University, Nanjing University, and Chinese University of Hong Kong. His current research interests are in the areas of mass estimation and mass-based approaches, ensemble approaches and data stream data mining. He is an associate editor for Journal of Data Mining and Knowledge Discovery. He co-chaired the Pacific-Asia Conference on Knowledge Discovery and Data Mining 2008. He has served as a member of program committees for a number of international conferences including ACM SIGKDD, IEEE ICDM and ICML. His research projects are supported by grants from Australian Research Council, US Air Force of Scientific Research (AFOSR/AOARD), and Australian Institute of Sport. Awards received include the Runner-up Best Paper Award in 2008 IEEE ICDM, and the Best Paper Award in 2006 PAKDD. He is the creator of isolation techniques and mass estimation.