Multiscale likelihood analysis and complexity penalized estimation.

DOI10.1214/009053604000000076MaRDI QIDQ1879941zbMATH OpenOpenAlexFDO

Authors Eric D. Kolaczyk, Robert D. Nowak

Publication date 15 September 2004

Published in The Annals of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/math/0406424

zbMATH Keywords

factorizationHaar basesHellinger distanceKullback-Leibler divergenceminimaxmodel selectionmultiresolutionrecursive partitioningthresholding estimators

Mathematics Subject Classification ID

Nontrigonometric harmonic analysis involving wavelets and other special systems (42C40) Probability distributions: general theory (60E05) Minimax procedures in statistical decision theory (62C20) Nonparametric estimation (62G05)

Abstract: We describe here a framework for a certain class of multiscale likelihood factorizations wherein, in analogy to a wavelet decomposition of an L^2 function, a given likelihood function has an alternative representation as a product of conditional densities reflecting information in both the data and the parameter vector localized in position and scale. The framework is developed as a set of sufficient conditions for the existence of such factorizations, formulated in analogy to those underlying a standard multiresolution analysis for wavelets, and hence can be viewed as a multiresolution analysis for likelihoods. We then consider the use of these factorizations in the task of nonparametric, complexity penalized likelihood estimation. We study the risk properties of certain thresholding and partitioning estimators, and demonstrate their adaptivity and near-optimality, in a minimax sense over a broad range of function spaces, based on squared Hellinger distance as a loss function. In particular, our results provide an illustration of how properties of classical wavelet-based estimators can be obtained in a single, unified framework that includes models for continuous, count and categorical data types.

Recommendations

• Multiscale statistical models
• Multiscale generalised linear models for nonparametric function estimation
• Multiscale Poisson Data Smoothing
• Multinomial Probability Estimation by Wavelet Thresholding
• Statistical models: conventional, penalized and hierarchical likelihood

Cites work

• A new class of unbalanced Haar wavelets that form an unconditional basis for \(L_ p\) on general measure spaces
• A statistical multiscale framework for Poisson inverse problems
• Bayesian Multiscale Models for Poisson Processes
• CART and best-ortho-basis: a connection
• scientific article; zbMATH DE number 3860199 (Why is no real title available?)
• scientific article; zbMATH DE number 3602484 (Why is no real title available?)
• scientific article; zbMATH DE number 1215245 (Why is no real title available?)
• scientific article; zbMATH DE number 739533 (Why is no real title available?)
• scientific article; zbMATH DE number 1134987 (Why is no real title available?)
• scientific article; zbMATH DE number 1405811 (Why is no real title available?)
• scientific article; zbMATH DE number 1405823 (Why is no real title available?)
• scientific article; zbMATH DE number 3277086 (Why is no real title available?)
• Minimax risk over hyperrectangles, and implications
• Minimum complexity density estimation
• Multiscale modeling and estimation of Poisson processes with application to photon-limited imaging
• Reproducibility and natural exponential families with power variance functions
• Risk bounds for model selection via penalization
• Ten Lectures on Wavelets
• The Lifting Scheme: A Construction of Second Generation Wavelets
• Unconditional bases are optimal bases for data compression and for statistical estimation

Cited in

(19)

• Estimating the intensity of a random measure by histogram type estimators
• Data-driven wavelet-Fisz methodology for nonparametric function estimation
• Plugin procedure in segmentation and application to hyperspectral image segmentation
• Poisson noise reduction with non-local PCA
• On the covariance properties of certain multiscale spatial processes
• Multiple testing approaches for removing background noise from images
• Bayesian Smoothing of Photon-Limited Images with Applications in Astronomy
• Poisson Noise Reduction with Higher-Order Natural Image Prior Model
• Multiply-rooted multiscale models for large-scale estimation
• scientific article; zbMATH DE number 1405811 (Why is no real title available?)
• Fisher Exact Scanning for Dependency
• Filtering the Maximum Likelihood for Multiscale Problems
• Multiscale statistical models
• Multiscale change point inference. With discussion and authors' reply
• Estimator selection with respect to Hellinger-type risks
• Boundary Detection Using a Bayesian Hierarchical Model for Multiscale Spatial Data
• Multiscale analysis of neural spike trains
• Poisson inverse problems
• Bayesian multiscale smoothing of Gaussian noised images

This page was built for publication: Multiscale likelihood analysis and complexity penalized estimation.

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1879941)