Intrinsic dimension estimation: Advances and open problems

By Francesco Camastra, Antonino Staiano. 2016

233 citations recorded by the source corpus.

The paper reviews state-of-the-art of the methods of Intrinsic Dimension (ID) Estimation.The paper defines the properties that an ideal ID estimator should have.The paper reviews, under the above mentioned framework, the major ID estimation methods underlining their advances and the open problems. Dimensionality reduction methods are preprocessing techniques used for coping with high dimensionality. They have the aim of projecting the original data set of dimensionality N, without information loss, onto a lower M-dimensional submanifold. Since the value of M is unknown, techniques that allow knowing in advance the value of M, called intrinsic dimension (ID), are quite useful. The aim of the paper is to review state-of-the-art of the methods of ID estimation, underlining the recent advances and the open problems.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Statistics and Probability, Representation & Self-Supervised Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Statistics and Probability

In Representation & Self-Supervised Learning

Citations in the corpus

Dataset Distillation with Infinitely Wide Convolutional Networks (2021, background): “ID can be defined and estimated differently based on assumptions on the manifold structure and the probability density function of the images on this manifold (see [Camastra and Staiano, 2016] for review).”
Learning algebraic varieties from samples (2018, methodology): “[12, 13]) on computing an intrinsic dimension of the sample Ω from a manifold V .”