Fast multivariate log-concave density estimation

DOI10.1016/J.CSDA.2019.04.005zbMATH Open1496.62021arXiv1805.07272OpenAlexW2963662588WikidataQ127760698 ScholiaQ127760698MaRDI QIDQ2337320FDOQ2337320

Authors: Fabian Rathke, Christoph Schnörr 

Publication date: 19 November 2019

Published in: Computational Statistics and Data Analysis (Search for Journal in Brave)

Abstract: A novel computational approach to log-concave density estimation is proposed. Previous approaches utilize the piecewise-affine parametrization of the density induced by the given sample set. The number of parameters as well as non-smooth subgradient-based convex optimization for determining the maximum likelihood density estimate cause long runtimes for dimensions $dgeq2$ and large sample sets. The presented approach is based on mildly non-convex smooth approximations of the objective function and extit{sparse}, adaptive piecewise-affine density parametrization. Established memory-efficient numerical optimization techniques enable to process larger data sets for dimensions $dgeq2$. While there is no guarantee that the algorithm returns the maximum likelihood estimate for every problem instance, we provide comprehensive numerical evidence that it does yield near-optimal results after significantly shorter runtimes. For example, 10000 samples in $mathbb{R}^2$ are processed in two seconds, rather than in $approx14$ hours required by the previous approach to terminate. For higher dimensions, density estimation becomes tractable as well: Processing $10000$ samples in $mathbb{R}^6$ requires 35 minutes. The software is publicly available as CRAN R package fmlogcondens.

Full work available at URL: https://arxiv.org/abs/1805.07272

Recommendations

• A computational approach to log-concave density estimation
• Maximum Likelihood Estimation of a Multi-Dimensional Log-Concave Density
• A fast algorithm for univariate log-concave density estimation
• Computing maximum likelihood estimators of a log-concave density function
• Adaptation in multivariate log-concave density estimation

zbMATH Keywords

maximum likelihood estimationnonparametric density estimationlog-concavityadaptive piecewise-affine parametrization

Mathematics Subject Classification ID

Computational methods for problems pertaining to statistics (62-08) Density estimation (62G07) Software, source code, etc. for problems pertaining to statistics (62-04)

Cites Work

• Title not available (Why is that?)
• On the estimation of a probability density function by the maximum penalized likelihood method
• Title not available (Why is that?)
• Maximum likelihood estimation of a log-concave density and its distribution function: basic properties and uniform consistency
• Sparse grids
• Maximum Likelihood Estimation of a Multi-Dimensional Log-Concave Density
• The quickhull algorithm for convex hulls
• The maximum numbers of faces of a convex polytope
• A modified BFGS method and its global convergence in nonconvex minimization
• Computational results of an \(O^{\ast }(n^{4})\) volume algorithm
• Invariant Integration Formulas for the n-Simplex by Combinatorial Methods
• The geometry of logconcave functions and sampling algorithms
• Higher-dimensional Voronoi diagrams in linear expected time
• Smoothed log-concave maximum likelihood estimation with applications
• Recent progress in log-concave density estimation
• Sample-optimal density estimation in nearly-linear time
• Hit-and-run for numerical integration

Cited In (11)

• High-dimensional nonparametric density estimation via symmetry and shape constraints
• Exact solutions in log-concave maximum likelihood estimation
• Nonparametric density estimation and bandwidth selection with B-spline bases: a novel Galerkin method
• A fast algorithm for univariate log-concave density estimation
• A new computational framework for log-concave density estimation
• Regularized multitask learning for multidimensional log-density gradient estimation
• Multivariate log-concave distributions as a nearly parametric model
• A computational approach to log-concave density estimation
• Shape Constrained Kernel PDF and PMF Estimation
• Density estimation for shift-invariant multidimensional distributions
• Assessing log-concavity of multivariate densities

Uses Software

• mclust
• R
• CRAN
• Qhull
• LogConcDEAD
• fmlogcondens

This page was built for publication: Fast multivariate log-concave density estimation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2337320)