Fourier-Information Duality in the Identity Management Problem
Abstract
We compare two recently proposed approaches for representing probability distributions over the space of permutations in the context of multi-target tracking. We show that these two representations, the Fourier approximation and the information form approximation can both be viewed as low dimensional projections of a true distribution, but with respect to different metrics. We identify the strengths and weaknesses of each approximation, and propose an algorithm for converting between the two forms, allowing for a hybrid approach that draws on the strengths of both representations. We show experimental evidence that there are situations where hybrid algorithms are favorable.
Cite
Text
Jiang et al. "Fourier-Information Duality in the Identity Management Problem." European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2011. doi:10.1007/978-3-642-23783-6_7Markdown
[Jiang et al. "Fourier-Information Duality in the Identity Management Problem." European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2011.](https://mlanthology.org/ecmlpkdd/2011/jiang2011ecmlpkdd-fourierinformation/) doi:10.1007/978-3-642-23783-6_7BibTeX
@inproceedings{jiang2011ecmlpkdd-fourierinformation,
title = {{Fourier-Information Duality in the Identity Management Problem}},
author = {Jiang, Xiaoye and Huang, Jonathan and Guibas, Leonidas J.},
booktitle = {European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases},
year = {2011},
pages = {97-113},
doi = {10.1007/978-3-642-23783-6_7},
url = {https://mlanthology.org/ecmlpkdd/2011/jiang2011ecmlpkdd-fourierinformation/}
}