scientific article; zbMATH DE number 17213

MaRDI QIDQ3973910zbMATH OpenFDO

Authors Antonio Cuevas, Wenceslao González-Manteiga

Publication date 26 June 1992

zbMATH Keywords

almost sure uniform consistencyautomatic choice of bandwidth parametersfixed finite number of inflection pointspiecewise convexitypseudo-likelihood cross-validation smoothing methodssimulationuniform convergenceunivariate kernel density estimatorsweak convergence

Mathematics Subject Classification ID

Density estimation (62G07) Asymptotic properties of nonparametric inference (62G20)

Review text

The authors establish a sufficient condition, in terms of piecewise convexity and weak convergence, for the uniform convergence of a sequence of densities with a fixed finite number of inflection points. This result is used to outline a new method for automatic choice of bandwidth (smoothing) parameters in univariate kernel density estimators. The resulting estimators are almost surely uniformly consistent. A relationship between the authors' results and Silverman's work on hypothesis testing on the number of modes is suggested. Finally, a simulation is used to compare the performance of the authors' inflection point method with that of ordinary pseudo-likelihood cross-validation smoothing methods. For the entire collection see [Zbl 0739.60024].

Recommendations

• The consistency of automatic kernel density estimates
• DATA-BASED CHOICE OF THE SMOOTHING PARAMETER FOR A KERNEL DENSITY ESTIMATOR
• Kernel contrasts: a data-based method of choosing smoothing parameters in nonparametric density estimation
• A comparative study of several smoothing methods in density estimation
• A method for the choice of smoothing parameter

Cited in

(11)

• Some asymptotics for multimodality tests based on kernel density estimates
• A comparative study of several smoothing methods in density estimation
• Bump hunting with non-Gaussian kernels
• Improving bandwidth selection methods by adding quantitative constraints
• A population background for nonparametric density-based clustering
• Signs of divided differences yield least squares data fitting with constrained monotonicity or convexity
• Identifiability of nonparametric mixture models and Bayes optimal clustering
• On the consistency of kernel density estimates under modality constraints
• Characterizations of Smooth Ambiguity Based on Continuous and Discrete Data
• Data Smoothing using Non-Negative Divided Differences and l₂ Approximation
• Minimum distance density-based estimation

This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3973910)