We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

math.ST

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Statistics Theory

Title: Fast adaptive estimation of log-additive exponential models in Kullback-Leibler divergence

Abstract: We study the problem of nonparametric estimation of density functions with a product form on the domain $\triangle=\{( x_1, \ldots, x_d)\in \mathbb{R}^d, 0\leq x_1\leq \dots \leq x_d \leq 1\}$. Such densities appear in the random truncation model as the joint density function of observations. They are also obtained as maximum entropy distributions of order statistics with given marginals. We propose an estimation method based on the approximation of the logarithm of the density by a carefully chosen family of basis functions. We show that the method achieves a fast convergence rate in probability with respect to the Kullback-Leibler divergence for densities whose logarithm belongs to a Sobolev function class with known regularity. In the case when the regularity is unknown, we propose an estimation procedure using convex aggregation of the log-densities to obtain adaptability. The performance of this method is illustrated in a simulation study.
Comments: 34 pages, 7 figures
Subjects: Statistics Theory (math.ST)
MSC classes: 62G07, 62G05, 62G20
Cite as: arXiv:1604.06304 [math.ST]
  (or arXiv:1604.06304v1 [math.ST] for this version)

Submission history

From: Richard Fischer [view email]
[v1] Thu, 21 Apr 2016 13:43:44 GMT (742kb,D)

Link back to: arXiv, form interface, contact.