Uniform spaces and the Newtonian structure of (big)data affinity kernels
General Topology
2017-01-16 v1 Analysis of PDEs
Abstract
Let be a (data) set. Let be a measure of the affinity between the data points and . We prove that has the structure of a Newtonian potential with decreasing and a quasi-metric on under two mild conditions on . The first is that the affinity of each to itself is infinite and that for the affinity is positive and finite. The second is a quantitative transitivity; if the affinity between and is larger than and the affinity of and is also larger than , then the affinity between and is larger than . The function is concave, increasing, continuous from onto with for every .
Keywords
Cite
@article{arxiv.1701.03746,
title = {Uniform spaces and the Newtonian structure of (big)data affinity kernels},
author = {Hugo Aimar and Ivana Gómez},
journal= {arXiv preprint arXiv:1701.03746},
year = {2017}
}
Comments
7 pages