Parameter estimation in systems exhibiting spatially complex solutions via persistent homology and machine learning

Research output: Contribution to journalArticlepeer-review

Abstract

We use persistent homology to extract topological information from complex spatio-temporal data generated by differential equations and use this information to estimate the corresponding parameters of the differential equation using regression methods in machine learning. We apply this technique to a predator–prey system and to the complex Ginzburg–Landau equation.

Original languageAmerican English
Pages (from-to)719-732
Number of pages14
JournalMathematics and Computers in Simulation
Volume185
DOIs
StatePublished - Jul 2021
Externally publishedYes

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science
  • Numerical Analysis
  • Modeling and Simulation
  • Applied Mathematics

Keywords

  • KNeighbors
  • Machine learning
  • Parameter estimation
  • Persistence diagrams
  • Persistent homology
  • SVR

Fingerprint

Dive into the research topics of 'Parameter estimation in systems exhibiting spatially complex solutions via persistent homology and machine learning'. Together they form a unique fingerprint.

Cite this