Local access to sparse connected subgraphs via edge sampling
Name
1192544421-MIT.pdf
Size
589.42 KB
Format
Adobe PDF
Checksum (MD5)
3aecde95338a8aa84452bb1f9f0d0f9e
Author(s)
Epstein, Rogers(Rogers S.)
Advisor(s)
Ronitt Rubinfeld.
Date Issued
2020
Publisher
Massachusetts Institute of Technology
Abstract
We contribute an approach to the problem of locally computing sparse connected subgraphs of dense graphs. In this setting, given an edge in a connected graph G = (V,E), an algorithm locally decides its membership in a sparse connected subgraph G* = (V,E*), where E* [subset] E and lE*l = o(lEl). Such an approach to subgraph construction is useful when dealing with massive graphs, where reading in the graph's full network description is impractical. While most prior results in this area require assumptions on G or that lE'l 0, we relax these assumptions. Given a general graph and a parameter T, we provide membership queries to a subgraph with O(lVlT) edges using Õ(lEl/T) probes. This is the first algorithm to work on general graphs and allow for a tradeoff between its probe complexity and the number of edges in the resulting subgraph. We achieve this result with ideas motivated from edge sparsification techniques that were previously unused in this problem. We believe these techniques will motivate new algorithms for this problem and related ones. Additionally, we describe an efficient method to access any node's neighbor set in a sparsified version of G where each edge is deleted with some i.i.d. probability.
Description
Thesis: M. Eng., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, May, 2020
Cataloged from the official PDF of thesis.
Includes bibliographical references (pages 37-38).
Subjects
Electrical Engineering and Computer Science.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
MIT theses may be protected by copyright. Please reuse MIT thesis content according to the MIT Libraries Permissions Policy, which is available through the URL provided.
Persistent DSpace Link
