On geodesic exponential kernels

Publikation: Bidrag til bog/antologi/rapportKonferenceabstrakt i proceedingsForskningfagfællebedømt

We consider kernel methods on general geodesic metric spaces and provide both negative and positive results. First we show that the common Gaussian kernel can only be generalized to a positive definite kernel on a geodesic metric space if the space is flat. As a result, for data on a Riemannian manifold, the geodesic Gaussian kernel is only positive definite if the Riemannian manifold is Euclidean. This implies that any attempt to design geodesic Gaussian kernels on curved Riemannian manifolds is futile. However, we show that for spaces with conditionally negative definite distances the geodesic Laplacian kernel can be generalized while retaining positive definiteness. This implies that geodesic Laplacian kernels can be generalized to some curved spaces, including spheres and hyperbolic spaces. Our theoretical results are verified empirically.

OriginalsprogEngelsk
TitelProceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
RedaktørerAasa Feragen, Marcello Pelillo, Marco Loog
Antal sider3
ForlagSpringer
Publikationsdato2015
Sider211-213
ISBN (Trykt)9781467369640
DOI
StatusUdgivet - 2015
Begivenhed3rd International Workshop on Similarity-Based Pattern Recognition, SIMBAD 2015 - Copenhagen, Danmark
Varighed: 12 okt. 201514 okt. 2015

Konference

Konference3rd International Workshop on Similarity-Based Pattern Recognition, SIMBAD 2015
LandDanmark
ByCopenhagen
Periode12/10/201514/10/2015

ID: 160634529