 |
infospec.sourceforge.net |
Information-theoretic derivation of min-cut based clustering
Infospec is software for comparing the long-time and short-time approximations to relevance information for fast mixing graphs. Relevance information is defined by the Information Bottleneck method applied to clustering on graphs as described in "Information-theoretic approach to network modularity". This software runs the experiments and generates the plots described in " Information-theoretic derivation of min-cut-based clustering". Infospec is implemented in MATLAB and can be downloaded at the sourceforge page.
This software is free for scientific use. Please contact us if you plan to use this software for commercial purposes. Do not further distribute without prior permission of the authors. If used in your scientific work, please cite as:
Anil Raj and Chris Wiggins, "An information-theoretic derivation of min-cut-based clustering",
IEEE Transactions of Pattern Analysis and Machine Intelligence, vol. 32 no. 6, pp. 988-995 (2010)
(
paper, bibtex citation)
Please forward links to related and resulting publications so we can learn about your research.
A screenshot of infospec:
)
Approximation errors of relevance information.
This work was funded by the National Institutes of Health under grants
NIH 1U54CA121852-01A1 and NIH 5PN2EY016586-03, and the National Science
Foundation under grant NSF IIS-0705580.