Login / Signup

Two efficient calculations of edit distance between marked point processes.

Yoshito HirataNoriyoshi Sukegawa
Published in: Chaos (Woodbury, N.Y.) (2019)
In this paper, we propose to use linear programming methods or a more specialized method, namely, the Hungarian method, for speeding up the exact calculation of an edit distance for marked point processes [Y. Hirata and K. Aihara, Chaos 25, 123117 (2015)]. The key observation is that the problem of calculating the edit distance reduces to a matching problem on a bipartite graph. Our preliminary numerical results show that the proposed implementations are faster than the conventional ones by a factor of 10-1000.
Keyphrases
  • density functional theory
  • molecular dynamics
  • palliative care
  • molecular dynamics simulations
  • convolutional neural network
  • machine learning