Manifold learning for parameter reduction

Alexander Holiday, Mahdi Kooshkbaghi, Juan M. Bello-Rivas, C. William Gear, Antonios Zagaris*, Ioannis G. Kevrekidis*

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

27 Citations (Scopus)


Large scale dynamical systems (e.g. many nonlinear coupled differential equations)can often be summarized in terms of only a few state variables (a few equations), a trait that reduces complexity and facilitates exploration of behavioral aspects of otherwise intractable models. High model dimensionality and complexity makes symbolic, pen–and–paper model reduction tedious and impractical, a difficulty addressed by recently developed frameworks that computerize reduction. Symbolic work has the benefit, however, of identifying both reduced state variables and parameter combinations that matter most (effective parameters, “inputs”); whereas current computational reduction schemes leave the parameter reduction aspect mostly unaddressed. As the interest in mapping out and optimizing complex input–output relations keeps growing, it becomes clear that combating the curse of dimensionality also requires efficient schemes for input space exploration and reduction. Here, we explore systematic, data-driven parameter reduction by means of effective parameter identification, starting from current nonlinear manifold-learning techniques enabling state space reduction. Our approach aspires to extend the data-driven determination of effective state variables with the data-driven discovery of effective model parameters, and thus to accelerate the exploration of high-dimensional parameter spaces associated with complex models.

Original languageEnglish
Pages (from-to)419-431
JournalJournal of computational physics
Publication statusPublished - Sept 2019


  • Data driven perturbation theory
  • Data mining
  • Diffusion maps
  • Model reduction
  • Parameter sloppiness


Dive into the research topics of 'Manifold learning for parameter reduction'. Together they form a unique fingerprint.

Cite this