Showing 1–1 of 1 results for author: Grillet, L

Effective Estimation of the Dimensions of a Manifold from Random Samples

Authors: Lucien Grillet, Juan Souto

Abstract: We give explicit theoretical and heuristical bounds for how big does a data set sampled from a reach-1 submanifold M of euclidian space need to be, to be able to estimate the dimension of M with 90% confidence. We give explicit theoretical and heuristical bounds for how big does a data set sampled from a reach-1 submanifold M of euclidian space need to be, to be able to estimate the dimension of M with 90% confidence. △ Less

Submitted 7 September, 2022; v1 submitted 5 September, 2022; originally announced September 2022.

Comments: 34 pages, 4 figures

ACM Class: I.3.5