Empirical survival Jensen-Shannon divergence as a goodness-of-fit measure for maximum likelihood estimation and curve fitting

DOI10.1080/03610918.2019.1630435MaRDI QIDQ5082812zbMATH OpenOpenAlexWikidataFDO

Authors M. Levene, Aleksejus Kononovicius

Publication date 21 June 2022

Published in Communications in Statistics. Simulation and Computation (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1809.11052, https://eprints.bbk.ac.uk/id/eprint/27719/1/27719.pdf

zbMATH Keywords

curve fittingdivergence measuresgoodness-of-fitmaximum likelihoodsurvival Jensen-Shannon divergence

Mathematics Subject Classification ID

Statistical aspects of information-theoretic topics (62B10) Point estimation (62F10) Measures of information, entropy (94A17)

Abstract: The coefficient of determination, known as $R^2$, is commonly used as a goodness-of-fit criterion for fitting linear models. $R^2$ is somewhat controversial when fitting nonlinear models, although it may be generalised on a case-by-case basis to deal with specific models such as the logistic model. Assume we are fitting a parametric distribution to a data set using, say, the maximum likelihood estimation method. A general approach to measure the goodness-of-fit of the fitted parameters, which is advocated herein, is to use a nonparametric measure for comparison between the empirical distribution, comprising the raw data, and the fitted model. In particular, for this purpose we put forward the Survival Jensen-Shannon divergence ($SJS$) and its empirical counterpart ($calESJS$) as a metric which is bounded, and is a natural generalisation of the Jensen-Shannon divergence. We demonstrate, via a straightforward procedure making use of the $calESJS$, that it can be used as part of maximum likelihood estimation or curve fitting as a measure of goodness-of-fit, including the construction of a confidence interval for the fitted parametric distribution. Furthermore, we show the validity of the proposed method with simulated data, and three empirical data sets.

Recommendations

• An \(R\)-squared measure of goodness of fit for some common nonlinear regression models
• The Jensen-Shannon divergence
• An \(R\)-squared measure of goodness of fit for exponential dispersion models
• scientific article; zbMATH DE number 2063764
• A Jensen-Gini measure of divergence with application in parameter estimation

Cites work

• A garden of integrals
• A new metric for probability distributions
• A Normal Approximation for Binomial, F, Beta, and Other Common, Related Tail Probabilities, I
• A widely applicable Bayesian information criterion
• An \(R\)-squared measure of goodness of fit for some common nonlinear regression models
• An analysis of edge detection by using the Jensen-Shannon divergence
• Bayes Factors
• Cumulative Residual Entropy: A New Measure of Information
• Divergence measures based on the Shannon entropy
• Elements of Information Theory
• Empirical analysis and agent-based modeling of the Lithuanian parliamentary elections
• Empirical properties of asset returns: stylized facts and statistical issues
• Essentials of econophysics modelling
• Fixed-effect variable selection in linear mixed models using \(R^2\) statistics
• Handbook of statistical distributions with applications
• scientific article; zbMATH DE number 3962966 (Why is no real title available?)
• scientific article; zbMATH DE number 1104922 (Why is no real title available?)
• scientific article; zbMATH DE number 2106909 (Why is no real title available?)
• Likelihood Ratio Tests for Model Selection and Non-Nested Hypotheses
• Model Selection and Multimodel Inference
• Nonparametric Estimation from Incomplete Observations
• Probability for statisticians
• Time series analysis: Methods and applications
• Tutorial on maximum likelihood estimation
• Unbiased Weibull modulus estimation using differential cumulative entropy

Describes a project that uses

Uses Software

• boot

This page was built for publication: Empirical survival Jensen-Shannon divergence as a goodness-of-fit measure for maximum likelihood estimation and curve fitting

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5082812)