A Generalization of Self-Improving Algorithms
Abstract
Ailon et al. [SICOMP'11] proposed self-improving algorithms for sorting and Delaunay triangulation (DT) when the input instances follow some unknown \emph{product distribution}. That is, comes from a fixed unknown distribution , and the 's are drawn independently. After spending time in a learning phase, the subsequent expected running time is , where , and and are the entropies of the distributions of the sorting and DT output, respectively. In this paper, we allow dependence among the 's under the \emph{group product distribution}. There is a hidden partition of into groups; the 's in the -th group are fixed unknown functions of the same hidden variable ; and the 's are drawn from an unknown product distribution. We describe self-improving algorithms for sorting and DT under this model when the functions that map to 's are well-behaved. After an -time training phase, we achieve and expected running times for sorting and DT, respectively, where is the inverse Ackermann function.
Keywords
Cite
@article{arxiv.2003.08329,
title = {A Generalization of Self-Improving Algorithms},
author = {Siu-Wing Cheng and Man-Kwun Chiu and Kai Jin and Man Ting Wong},
journal= {arXiv preprint arXiv:2003.08329},
year = {2020}
}