PreDeCon

This repository is not associated with the original authors of [Boehm,2004].

About

Subspace Preference Weighted Density Connected Clustering (PreDeCon) [Boehm,2004] can be seen as a modification to the famous DBSCAN [Ester,1996] that addresses problems which arise in high-dimensional spaces.

Installation

Install with pip.

From PyPI

$ pip install predecon-exioreed

Alternatively, from source

$ pip install git+https://github.com/exioReed/PreDeCon@master#egg=PreDeCon-exioreed

or

$ git clone https://github.com/exioReed/PreDeCon.git
$ cd PreDeCon
$ pip install .

References

[Boehm,2004] Boehm, C. et al., "Density Connected Clustering with Local Subspace Preferences". In: Proceedings of the 4th IEEE Internation Conference on Data Mining (ICDM), Brighton, UK, 2004.

[Ester,1996] Ester, M. et al., "A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise". In: Proceedings of the 2nd International Conference on Knowledge Discovery and Data Mining, Portland, OR, 1996.