This repository contains implementations of the Random Cut Forest (RCF) probabilistic data structure. RCFs were originally developed at Amazon to use in a nonparametric anomaly detection algorithm for streaming data. Later new algorithms based on RCFs were developed for density estimation, imputation, and forecasting.
The different directories correspond to equivalent implementations in different languages, and bindings to to those base implementations, using language specific features for greater flexibility of use.
RandomCutForest in the randomcutforest-core package provides an estimation (say anomaly score, or extrapolation over a forecast horizon) and using that raw estimation can be challenging. The randomcutforest-parkservices package provides several capabilities (ThresholdedRandomCutForest, RCFCaster, respectively) for distilling the scores to a determination of a potential anomaly or calibrated forecast respectively. The package randomcutforest-examples showcases several example scenarios for using the repository. They also provide examples for some of the parameter settings. Many of these examples are built in tests.
- Guha, S., Mishra, N., Roy, G., & Schrijvers, O. (2016, June). Robust random cut forest based anomaly detection on streams. In International conference on machine learning (pp. 2712-2721).
This project has adopted an Open Source Code of Conduct.
If you discover a potential security issue in this project we ask that you notify AWS/Amazon Security via our vulnerability reporting page. Please do not create a public GitHub issue.
See the LICENSE file for our project's licensing. We will ask you to confirm the licensing of your contribution.
Copyright 2019-2020 Amazon.com, Inc. or its affiliates. All Rights Reserved.