eduzhai > Applied Sciences > Engineering >

Graph signal processing for machine learning A review and new perspectives

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 22 pages

Abstract: The effective representation, processing, analysis, and visualization oflarge-scale structured data, especially those related to complex domains suchas networks and graphs, are one of the key questions in modern machinelearning. Graph signal processing (GSP), a vibrant branch of signal processingmodels and algorithms that aims at handling data supported on graphs, opens newpaths of research to address this challenge. In this article, we review a fewimportant contributions made by GSP concepts and tools, such as graph filtersand transforms, to the development of novel machine learning algorithms. Inparticular, our discussion focuses on the following three aspects: exploitingdata structure and relational priors, improving data and computationalefficiency, and enhancing model interpretability. Furthermore, we provide newperspectives on future development of GSP techniques that may serve as a bridgebetween applied mathematics and signal processing on one side, and machinelearning and network science on the other. Cross-fertilization across thesedifferent disciplines may help unlock the numerous challenges of complex dataanalysis in the modern age.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×