A quantum Jensen-Shannon graph kernel using discrete-time quantum walks


In this paper, we develop a new graph kernel by using the quantum Jensen-Shannon divergence and the discrete-time quantum walk. To this end, we commence by performing a discrete-time quantum walk to compute a density matrix over each graph being compared. For a pair of graphs, we compare the mixed quantum states represented by their density matrices using the quantum Jensen-Shannon divergence. With the density matrices for a pair of graphs to hand, the quantum graph kernel between the pair of graphs is defined by exponentiating the negative quantum Jensen-Shannon divergence between the graph density matrices. We evaluate the performance of our kernel on several standard graph datasets, and demonstrate the effectiveness of the new kernel.

Publication DOI: https://doi.org/10.1007/978-3-319-18224-7_25
Divisions: College of Engineering & Physical Sciences
College of Engineering & Physical Sciences > Systems analytics research institute (SARI)
Additional Information: The final publication is available at Springer via http://dx.doi.org/10.1007/978-3-319-18224-7_25 Funding: UK Royal Society
Event Title: 10th IAPR-TC-15 international workshop, GbRPR 2015
Event Type: Other
Event Dates: 2015-05-13 - 2015-05-15
ISBN: 978-3-319-18223-0, 978-3-319-18224-7
Full Text Link: http://link.spr ... -319-18224-7_25
Related URLs:
PURE Output Type: Conference contribution
Published Date: 2015
Authors: Bai, Lu
Rossi, Luca (ORCID Profile 0000-0002-6116-9761)
Ren, Peng
Zhang, Zhihong
Hancock, Edwin R.



Version: Accepted Version

Export / Share Citation


Additional statistics for this record