Consensus Ranking with Signed Permutations
Abstract
Signed permutations (also known as the hyperoctahedral group) are used in modeling genome rearrangements. The algorithmic problems they raise are computationally demanding when not NP-hard. This paper presents a tractable algorithm for learning consensus ranking between signed permutations under the inversion distance. This can be extended to estimate a natural class of exponential models over the group of signed permutations. We investigate experimentally the efficiency of our algorithm for modeling data generated by random reversals.
Cite
Text
Arora and Meila. "Consensus Ranking with Signed Permutations." International Conference on Artificial Intelligence and Statistics, 2013.Markdown
[Arora and Meila. "Consensus Ranking with Signed Permutations." International Conference on Artificial Intelligence and Statistics, 2013.](https://mlanthology.org/aistats/2013/arora2013aistats-consensus/)BibTeX
@inproceedings{arora2013aistats-consensus,
title = {{Consensus Ranking with Signed Permutations}},
author = {Arora, Raman and Meila, Marina},
booktitle = {International Conference on Artificial Intelligence and Statistics},
year = {2013},
pages = {117-125},
url = {https://mlanthology.org/aistats/2013/arora2013aistats-consensus/}
}