Flow-Based Dissimilarities: Shortest Path, Commute Time, Max-Flow and Free Energy

被引：9

作者：

Guex, Guillaume ^{[1
]}

Bavaud, Francois ^{[1
]}

机构：

[1] Univ Lausanne, CH-1015 Lausanne, Switzerland

来源：

DATA SCIENCE, LEARNING BY LATENT STRUCTURES, AND KNOWLEDGE DISCOVERY | 2015年

关键词：

DISTANCES;

D O I：

10.1007/978-3-662-44983-7_9

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

Random-walk based dissimilarities on weighted networks have demonstrated their efficiency in clustering algorithms. This contribution considers a few alternative network dissimilarities, among which a new max-flow dissimilarity, and more general flow-based dissimilarities, freely mixing shortest paths and random walks in function of a free parameter-the temperature. Their geometrical properties and in particular their squared Euclidean nature are investigated through their power indices and multidimensional scaling properties. In particular, formal and numerical studies demonstrate the existence of critical temperatures, where flow-based dissimilarities cease to be squared Euclidean. The clustering potential of medium range temperatures is emphasized.

引用

页码：101 / 111

页数：11