We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Machine Learning in Network Centrality Measures: Tutorial and Outlook

Abstract: Complex networks are ubiquitous to several Computer Science domains. Centrality measures are an important analysis mechanism to uncover vital elements of complex networks. However, these metrics have high computational costs and requirements that hinder their applications in large real-world networks. In this tutorial, we explain how the use of neural network learning algorithms can render the application of the metrics in complex networks of arbitrary size. Moreover, the tutorial describes how to identify the best configuration for neural network training and learning such for tasks, besides presenting an easy way to generate and acquire training data. We do so by means of a general methodology, using complex network models adaptable to any application. We show that a regression model generated by the neural network successfully approximates the metric values and therefore are a robust, effective alternative in real-world applications. The methodology and proposed machine learning model use only a fraction of time with respect to other approximation algorithms, which is crucial in complex network applications.
Comments: 7 tables, 9 figures, version accepted at ACM Computing Surveys. this https URL
Subjects: Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE); Networking and Internet Architecture (cs.NI); Social and Information Networks (cs.SI); Machine Learning (stat.ML)
ACM classes: I.2.6, C.2, F.1.1
Journal reference: ACM Comput. Surv. 51, 5, Article 102 (October 2018), 32 pages
DOI: 10.1145/3237192
Cite as: arXiv:1810.11760 [cs.LG]
  (or arXiv:1810.11760v1 [cs.LG] for this version)

Submission history

From: Luis Lamb [view email]
[v1] Sun, 28 Oct 2018 04:51:08 GMT (2078kb)

Link back to: arXiv, form interface, contact.