Isolation Forest Variants

This repository contains reproducible code of ablation study to research randomization in isolation forest anomaly detection algorithm.

Currently, the repository contains the following isolation forest variants:

• Isolation Forest (3 original implementations which are mine, sklearn's, and pyod's.)
• Extended Isolation Forest (See citations section below.)
• Isolation Forest which samples an attribute value instead of choosing randomly as in original version.

Requirements

• Anaconda3

Setup

• conda env create -f environment.yml
• conda activate iforest-variants
• python experiments.py

Citations

If you use iForest, please consider using following references:

@inproceedings{liu2008isolation,
  title={Isolation forest},
  author={Liu, Fei Tony and Ting, Kai Ming and Zhou, Zhi-Hua},
  booktitle={2008 Eighth IEEE International Conference on Data Mining},
  pages={413--422},
  year={2008},
  organization={IEEE}
}

@article{liu2012isolation,
  title={Isolation-based anomaly detection},
  author={Liu, Fei Tony and Ting, Kai Ming and Zhou, Zhi-Hua},
  journal={ACM Transactions on Knowledge Discovery from Data (TKDD)},
  volume={6},
  number={1},
  pages={3},
  year={2012},
  publisher={Acm}
}

If you use extended iForest, please consider using following reference:

@ARTICLE{2018arXiv181102141H,
   author = {{Hariri}, S. and {Carrasco Kind}, M. and {Brunner}, R.~J.},
    title = "{Extended Isolation Forest}",
  journal = {ArXiv e-prints},
archivePrefix = "arXiv",
   eprint = {1811.02141},
 keywords = {Computer Science - Machine Learning, Statistics - Machine Learning},
     year = 2018,
    month = nov,
   adsurl = {http://adsabs.harvard.edu/abs/2018arXiv181102141H},
  adsnote = {Provided by the SAO/NASA Astrophysics Data System}
}

If you use scikit-learn or PyOD implementation of iForest, please consider using following references:

@article{pedregosa2011scikit,
  title={Scikit-learn: Machine learning in Python},
  author={Pedregosa, Fabian and Varoquaux, Ga{\"e}l and Gramfort, Alexandre and Michel, Vincent and Thirion, Bertrand and Grisel, Olivier and Blondel, Mathieu and Prettenhofer, Peter and Weiss, Ron and Dubourg, Vincent and others},
  journal={Journal of machine learning research},
  volume={12},
  number={Oct},
  pages={2825--2830},
  year={2011}
}

@article{zhao2019pyod,
  title={PyOD: A python toolbox for scalable outlier detection},
  author={Zhao, Yue and Nasrullah, Zain and Li, Zheng},
  journal={arXiv preprint arXiv:1901.01588},
  year={2019}
}