SMS scnews item created by Lachlan Smith at Mon 22 Jun 2020 1315
Type: Seminar
Distribution: World
Expiry: 17 Aug 2020
Calendar1: 26 Jun 2020 1600-1700
CalLoc1: Zoom
CalTitle1: Sydney Dynamics Group: Caroline Wormell -- Spectral convergence of diffusion maps
Auth: (lsmi9789) in SMS-WASM

Sydney Dynamics Group: Wormell -- Spectral convergence of diffusion maps

Dear all, 

This week, Friday June 26, Caroline Wormell (USyd) will give a talk at 4pm (Sydney time)
via Zoom.  

Zoom link: 

Meeting ID: 978 8608 0891 

Title: Spectral convergence of diffusion maps 

Diffusion maps is a manifold learning algorithm widely used for dimensionality
reduction.  Using a sample from a distribution, it approximates the eigenvalues and
eigenfunctions of associated Laplace-Beltrami operators.  Theoretical bounds on the
approximation error are however generally much weaker than the rates that are seen in
practice.  We present new approaches to improve the error bounds in the model case where
the distribution is supported on a hypertorus.  For the data sampling (variance)
component of the error we make spatially localised compact embedding estimates on
certain Hardy spaces; we study the deterministic (bias) component as a perturbation of
the Laplace-Beltrami operator’s associated PDE, and apply relevant spectral stability
results.  These techniques enable long-standing pointwise error bounds to be matched for
both the spectral data and the norm convergence of the operator discretisation.  

We also introduce an alternative normalisation for diffusion maps based on Sinkhorn
weights.  This normalisation approximates a Langevin diffusion on the sample and yields
a symmetric operator approximation.  We prove that it has better convergence compared
with the standard normalisation on flat domains, and present a highly efficient
algorithm to compute the Sinkhorn weights.  

Past talks can be found on the YouTube channel: 

I hope to see you all online.  


ball Calendar (ICS file) download, for import into your favourite calendar application
ball UNCLUTTER for printing
ball AUTHENTICATE to mark the scnews item as read
School members may try to .