11
15

High-dimensional nonparametric density estimation via symmetry and shape constraints

Abstract

We tackle the problem of high-dimensional nonparametric density estimation by taking the class of log-concave densities on Rp\mathbb{R}^p and incorporating within it symmetry assumptions, which facilitate scalable estimation algorithms and can mitigate the curse of dimensionality. Our main symmetry assumption is that the super-level sets of the density are KK-homothetic (i.e. scalar multiples of a convex body KRpK \subseteq \mathbb{R}^p). When KK is known, we prove that the KK-homothetic log-concave maximum likelihood estimator based on nn independent observations from such a density has a worst-case risk bound with respect to, e.g., squared Hellinger loss, of O(n4/5)O(n^{-4/5}), independent of pp. Moreover, we show that the estimator is adaptive in the sense that if the data generating density admits a special form, then a nearly parametric rate may be attained. We also provide worst-case and adaptive risk bounds in cases where KK is only known up to a positive definite transformation, and where it is completely unknown and must be estimated nonparametrically. Our estimation algorithms are fast even when nn and pp are on the order of hundreds of thousands, and we illustrate the strong finite-sample performance of our methods on simulated data.

View on arXiv
Comments on this paper