Frederic Chazal and Marina Meila
Announcements the workshop schedule will reflect further changes and updates to the program, if any.
Visit also the FORUM on Geometric Machine Learning and leave a comment. The organizers will consider these comments in preparation for the panel discussions that will take place during the workshop.
And, if you are at NIPS 2017, visit us in Room 103C.
This workshop will bring together researchers from the various subdisciplines of Geometric Data Analysis, such as manifold learning, topological data analysis, shape analysis, and will showcase recent progress in this field. The focus will be on high dimensional and big data, and on mathematically founded methodology.
One aim of this workshop is to build connections between Topological Data Analysis on one side and Manifold Learning on the other. This is starting to happen, after years of more or less separate evolution of the two fields. The workshop will accelerate this process.
The second aim is to bring GDA closer to real applications. This is necessary for manifold learning more than for TDA, since ML is still somehow disconnected from real problems (which shouldn't be the case), while TDA is not. Providing a view of real problems and real data will also be a motivator for people in GDA to see TDA and ML as one.
The impact of GDA in practice also depends on having scalable implementations of the most current results in theory. This workshop will showcase the GDA tools which achieve this.
This will be a TWO day workshop containing invited talks, contributed talks, a poster session and period and a (panel) discussion session.
We invite and will select submission from all several areas: mathematics, statistics, computer science, as well as from groups of scientists who work with GDA.
The submissions will be of two categories original contributions and (NIPS) refresh (work published recently). We deem this second section useful in order to build context for the discussions and contributions to be made in
We invite and contributions from several areas: mathematics, statistics, computer science, as well as from groups of scientists who work with GDA.
The submissions will be of two categories original contributions and (NIPS) refresh (work published recently). We deem this second section useful in order to build context for the discussions and contributions to be made in the workshop.
Thanks to all presenters who made their slides available!
Fri Dec 08 08:10 AM - 09:10 AM Invited talk Supervised learning of labeled pointcloud differences via cover-tree entropy reduction Harer
Fri Dec 08 09:10 AM - 09:40 AM Estimating the Reach of a Manifold Aamari
Fri Dec 08 09:40 AM - 10:10 AM Multiscale geometric feature extraction Polonik
Fri Dec 08 10:10 AM - 10:30 AM Poster spotlights
On The Information Geometry of Word Embedding Riccardo Volpi, D. Marinelli, P. Hlihor, and L. Malago
Parallel multi-scale reduction of persistent homology Mendoza Smith
Maximum likelihood estimation of Riemannian metrics from Euclidean data Arvanitidis
A dual framework for low rank tensor completion Nimishakavi
Fri Dec 08 11:00 AM - 11:30 AM Persistent homology of KDE filtration of Rips complexes Shin, Rinaldo
Fri Dec 08 11:30 AM - 11:55 AM Characterizing non-linear dimensionality reduction methods using Laplacian-like operators Ting
Fri Dec 08 02:00 PM - 03:00 PM Invited talk Multiscale characterization of molecular dynamics Clementi
Fri Dec 08 03:30 PM - 04:00 PM Functional Data Analysis using a Topological Summary Statistic: the Smooth Euler Characteristic Transform Crawford
Fri Dec 08 04:00 PM - 04:30 PM Consistent manifold representation for TDA Sauer
Fri Dec 08 05:00 PM - 06:00 PM Discussion: Geometric Data Analysis
Sat Dec 09 08:30 AM - 08:50 AM Tutorial Topological Data Analisys with GUDHI and Scalable manifold learning and clustering with megaman Rouvreau, Meila
Sat Dec 09 08:50 AM - 09:20 AM Tutorial Introduction to the R package TDA Jisu Kim
Sat Dec 09 09:20 AM - 09:50 AM Riemannian metric estimation and the problem of isometric embedding Meila
Sat Dec 09 09:50 AM - 10:20 AM Invited talk Ordinal distance comparisons: from topology to geometry von Luxburg
Sat Dec 09 10:50 AM - 11:30 AM Discussion Geometric Data Analysis software
Sat Dec 09 02:00 PM - 02:30 PM Modal-sets, and density-based Clustering Kpotufe
Sat Dec 09 02:30 PM - 03:00 PM Community Trees in Networks Chen
Sat Dec 09 03:30 PM - 04:00 PM Beyond Two-sample-tests: Localizing Data Discrepancies in High-dimensional Spaces Cazals
For ORIGINAL track, please submit an extended abstract in NIPS format, no more than 5 pages long, plus an additional page of references. Alternatively, you can submit a link to an arxiv paper. Any work submitted here must be PREVIOUSLY UNPUBLISHED at the time of the NIPS workshop (work submitted but not yet published is acceptable).
For the REFRESH track, we accept work published in 2017 in conferences. Please submit a link to the conference proceedings and the pdf of the paper. If your submission contains new work that augments work already published, please tell us which parts are new (in the email that accompanies your paper). These submissions will be considered for the ORIGINAL track.
Send any submission to the workshop organizers, specifying if it is for the ORIGINAL or REFRESH track.
Marina Meila and Fred Chazal