Search Results for author: Vladimir G. Pestov

Found 3 papers, 0 papers with code

A learning problem whose consistency is equivalent to the non-existence of real-valued measurable cardinals

no code implementations4 May 2020 Vladimir G. Pestov

We show that the $k$-nearest neighbour learning rule is universally consistent in a metric space $X$ if and only if it is universally consistent in every separable subspace of $X$ and the density of $X$ is less than every real-measurable cardinal.

Universal consistency of the $k$-NN rule in metric spaces and Nagata dimension

no code implementations28 Feb 2020 Benoît Collins, Sushma Kumari, Vladimir G. Pestov

The generalization is non-trivial because of the distance ties being more prevalent in the non-euclidean setting, and on the way we investigate the relevant geometric properties of the metrics and the limitations of the Stone argument, by constructing various examples.

Elementos da teoria de aprendizagem de máquina supervisionada

no code implementations6 Oct 2019 Vladimir G. Pestov

The topics include: the geometry of the Hamming cube, concentration of measure, shattering and VC dimension, Glivenko-Cantelli classes, PAC learnability, universal consistency and the k-NN classifier in metric spaces, dimensionality reduction, universal approximation, sample compression.

Dimensionality Reduction

Cannot find the paper you are looking for? You can Submit a new open access paper.