Manifold learning and the quantum Jensen-Shannon divergence kernel

Luca Rossi, Andrea Torsello, Edwin R. Hancock

Research output: Chapter in Book/Published conference outputConference publication


The quantum Jensen-Shannon divergence kernel [1] was recently introduced in the context of unattributed graphs where it was shown to outperform several commonly used alternatives. In this paper, we study the separability properties of this kernel and we propose a way to compute a low-dimensional kernel embedding where the separation of the different classes is enhanced. The idea stems from the observation that the multidimensional scaling embeddings on this kernel show a strong horseshoe shape distribution, a pattern which is known to arise when long range distances are not estimated accurately. Here we propose to use Isomap to embed the graphs using only local distance information onto a new vectorial space with a higher class separability. The experimental evaluation shows the effectiveness of the proposed approach.

Original languageEnglish
Title of host publicationComputer Analysis of Images and Patterns
Subtitle of host publication15th international conference, CAIP 2013, York, UK, August 27-29, 2013, Proceedings
EditorsRichard Wilson, Edwin Hancock, Adrian Bors, William Smith
Place of PublicationBerlin (US)
Number of pages8
ISBN (Electronic)978-3-642-40261-6
ISBN (Print)978-3-642-40260-9
Publication statusPublished - 2013
Event15th international conference on Computer Analysis of Images and Patterns - York, United Kingdom
Duration: 27 Aug 201329 Aug 2013

Publication series

NameLecture notes in computer science
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference15th international conference on Computer Analysis of Images and Patterns
Abbreviated titleCAIP 2013
Country/TerritoryUnited Kingdom


  • continuous-time quantum walk
  • graph kernels
  • Manifold learning
  • quantum Jensen-Shannon divergence


Dive into the research topics of 'Manifold learning and the quantum Jensen-Shannon divergence kernel'. Together they form a unique fingerprint.

Cite this