A library for implementing the global k-means and the global k-means++ clustering algorithms.
Project description
The Global $k$-mean++ clustering algorithm
The global $k$-means++ is an effective relaxation of the global $k$-means clustering algorithm, providing an ideal compromise between clustering error and execution speed. It is an effective way of acquiring quality clustering solutions akin to those of global $k$-means with a reduced computational load. It is an incremental clustering approach that dynamically adds one cluster center at each $k$ cluster sub-problem. For each $k$ cluster sub-problem, the method selects $L$ data points as candidates for the initial position of the new center using the effective $k$-means++ selection probability distribution. The selection method is fast and requires no extra computational effort for distance computations.
@article{vardakas2022global,
title={Global $k$-means$++$: an effective relaxation of the global $k$-means clustering algorithm},
author={Vardakas, Georgios and Likas, Aristidis},
journal={arXiv preprint arXiv:2211.12271},
year={2022}
}
@article{likas2003global,
title={The global k-means clustering algorithm},
author={Likas, Aristidis and Vlassis, Nikos and Verbeek, Jakob J},
journal={Pattern recognition},
volume={36},
number={2},
pages={451--461},
year={2003},
publisher={Elsevier}
}
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for global_kmeans_pp-0.1.0-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | f2d14e1e40e71b847aeaeacb4a429a473eec2d0584bdef44a311c8235127e459 |
|
MD5 | f23defa8ca33e84be43603c8be940570 |
|
BLAKE2b-256 | e1cbabef54c38bccdc4160eca7a29e0f8664227e6682686fec13f881ab7d6a74 |