In this paper we present the Random Cut Forest algorithm, which detects anomalies in real-time streaming data. We have implemented this algorithm as a built-in SQL function in Amazon Kinesis Data Analytics, which is a fully managed AWS service that makes it easy to analyze streaming data with SQL in real-time.
This paper was published in the Proceeedings of the 33rd International Conference on Machine Learning, New York, NY, USA, 2016. JMLR: W&CP volume 48. Copyright 2016 by the author(s).