Intrinsic Dimension Estimation Using Packing Numbers

Part of Advances in Neural Information Processing Systems 15 (NIPS 2002)

Bibtex Metadata Paper


Balázs Kégl


We propose a new algorithm to estimate the intrinsic dimension of data sets. The method is based on geometric properties of the data and re- quires neither parametric assumptions on the data generating model nor input parameters to set. The method is compared to a similar, widely- used algorithm from the same family of geometric techniques. Experi- ments show that our method is more robust in terms of the data generating distribution and more reliable in the presence of noise.