Adaptive Sampling for Nonlinear Dimensionality Reduction Based on Manifold Learning

Thomas Franz, Ralf Zimmermann, Stefan Goertz

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

392 Downloads (Pure)


We make use of the non-intrusive dimensionality reduction method Isomap in order to emulate nonlinear parametric flow problems that are governed by the Reynolds-averaged Navier-Stokes equations. Isomap is a manifold learning approach that provides a low-dimensional embedding space that is approximately isometric to the manifold that is assumed to be formed by the high-fidelity Navier-Stokes flow solutions under smooth variations of the inflow conditions. The focus of the work at hand is the adaptive construction and refinement of the Isomap emulator: We exploit the non-Euclidean Isomap metric to detect and fill up gaps in the sampling in the embedding space. The performance of the proposed manifold filling method will be illustrated by numerical experiments, where we consider nonlinear parameter-dependent steady-state Navier-Stokes flows in the transonic regime.
Original languageEnglish
Title of host publicationModel Reduction of Parametrized Systems
EditorsPeter Brenner, Mario Ohlberger, Anthony Patera, Gianluigi Rozza, Karsten Urban
Publication date2017
ISBN (Print)978-3-319-58785-1
ISBN (Electronic)978-3-319-58786-8
Publication statusPublished - 2017
EventModel Reduction of Parametrized Systems III - SISSA, Trieste, Italy
Duration: 13. Oct 201516. Oct 2015


ConferenceModel Reduction of Parametrized Systems III
Internet address
SeriesModeling, Simulation & Applications

Fingerprint Dive into the research topics of 'Adaptive Sampling for Nonlinear Dimensionality Reduction Based on Manifold Learning'. Together they form a unique fingerprint.

Cite this