IsoKernel

IsoKernel - IsoKernel is a python library for Isolation Kernel. It includes several kernel methods using isolation mechanism.

Supported Kernel methods

• Isolation Kernel (IsoKernel)
• Isolation Distribution Kernel (IsoDisKernel)

Installing

PyPI install, presuming you have an up to date pip.

For a manual install of the latest code directly from GitHub:

pip install git+https://github.com/xhan97/IsoKernel.git

Alternatively download the package, install requirements, and manually run the installer:

wget https://codeload.github.com/xhan97/IsoKernel/zip/refs/heads/master
unzip IsoKernel-master.zip
rm IsoKernel-master.zip
cd IsoKernel-master

pip install -r requirements.txt

python setup.py install

How to use IsoKernel

The IsoKernel package inherits from sklearn classes, and thus drops in neatly next to other sklearn with an identical calling API. Similarly it supports input in a variety of formats: an array (or pandas dataframe) of shape (num_samples x num_features).

from IsoKernel import IsoKernel
from sklearn.datasets import make_blobs

data, _ = make_blobs(1000)

ik = IsoKernel(n_estimators=200, max_samples=16, method="anne") # method can be "anne" or "inne"
ik = ik.fit(data)
# get Isolation Kernel feature for all points in data.
ik.transform(data)
# get pairwise Isolation Kernel similarity for all points in data.
ik.similarity(data)

How to use IsoDisKernel

Isolation Distributional Kernel is a new way to measure the similarity between two distributions. It addresses two key issues of kernel mean embedding, where the kernel employed has:

1. a feature map with intractable dimensionality which leads to high computational cost;
2. data independency which leads to poor accuracy.

from IsoKernel import IsoDisKernel
from sklearn.datasets import make_blobs

data, _ = make_blobs(1000)

idk = IsoDisKernel(n_estimators=200, max_samples=16, method="anne") # method can be "anne" or "inne"
idk = idk.fit(data)
D_i = data[:10]
D_j = data[-10:]
# Directly get the similarity between two distributions
# is_normalize: whether return the normalized similarity matrix ranged of [0,1]. Default: True
sim = idk.similarity(D_i, D_j, is_normalize=True)

# get ik feature of two distributions
ikm_D_i, ikm_D_j = idk.transform(D_i, D_j)
# get kernel mean embedding
kme_D_i = idk.kernel_mean_embedding(ikm_D_i)
kme_D_j = idk.kernel_mean_embedding(ikm_D_j)
# get similarity between two distributions.
sim = idk.kme_similarity(kme_D_i, kme_D_j, is_normalize=True)

Running the Tests

The package tests can be run after installation using the command:

pip install pytest

or, if pytest is installed:

pytest  IsoKernel/tests

If one or more of the tests fail, please report a bug at https://github.com/xhan97/IsoKernel/issues

Python Version

Python 3 is recommend the better option if it is available to you.

Citing

If you have used this codebase in a scientific publication and wish to cite it, please use the following publication (Bibtex format):

@inproceedings{10.1145/3219819.3219990,
     author = {Ting, Kai Ming and Zhu, Yue and Zhou, Zhi-Hua},
     title = {Isolation Kernel and Its Effect on SVM},
     year = {2018},
     publisher = {Association for Computing Machinery},
     address = {New York, NY, USA},
     booktitle = {Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining},
     pages = {2329–2337},
     numpages = {9},
     location = {London, United Kingdom},
     series = {KDD '18}
     }

@inproceedings{ting2020Isolation,
    author = {Ting, Kai Ming and Xu, Bi-Cun and Washio, Takashi and Zhou, Zhi-Hua},
    title = {Isolation Distributional Kernel: A New Tool for Kernel Based Anomaly Detection},
    year = {2020},
    publisher = {Association for Computing Machinery},
    address = {New York, NY, USA},
    doi = {10.1145/3394486.3403062},
    pages = {198-206},
    numpages = {9},
    series = {KDD '20}
}

@inproceedings{HZTZL22Streaming,
 author = {Han, Xin and Zhu, Ye and Ting, Kai Ming and Zhan, De-Chuan and Li, Gang},
 title = {Streaming Hierarchical Clustering Based on Point-Set Kernel},
 year = {2022},
 isbn = {9781450393850},
 publisher = {Association for Computing Machinery},
 address = {New York, NY, USA},
 url = {https://doi.org/10.1145/3534678.3539323},
 doi = {10.1145/3534678.3539323},
 pages = {525–533},
 numpages = {9},
 keywords = {streaming data, hierarchical clustering, isolation kernel},
 location = {Washington DC, USA},
 series = {KDD '22}

}

License

Apache license