Asymptotic biases of information and cross-validation criteria under canonical parametrization

Ogasawara, Haruhiko

doi:10.1080/03610926.2017.1422759

Cited by 1 publication

(1 citation statement)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Model selection using AIC (Akaike's Information Criterion) ( [1], [2], Chapter 3 in [3],Chapter 2 and Chapter 7 in [4], Chapter 2 in [5]) coincides asymptotically with selection using cross-validation ( [6], [7], [8], [9], [10], [11]); a more recent and sophisticated study on this matter is [12]. However, this is an asymptotic result and is based on several assumptions; when dealing with practical data analysis, it is often not clear whether using AIC is the same as using cross-validation.…”

Section: Introductionmentioning

confidence: 99%

A Simulation Study for the AIC and Likelihood Cross-validation: The Case of Exponential Versus Weibull Distributions

Takezawa

2018

JAMCS

View full text Add to dashboard Cite

Various methods are available for choosing statistical models. It is difficult to know which model selection criterion is the best for specific data. This paper discusses a method for choosing the model selection criterion based on the characteristics of the data and models. As an example, we examined the choice between AIC and likelihood cross-validation as the model selection criterion with the exponential distribution and Weibull distribution as candidate models. First, we examined the characteristics of AIC and likelihood cross-validation using data generated from an exponential distribution or Weibull distribution; AIC and likelihood cross-validation show substantially different natures. Next, from the results of the numerical simulations, we propose an intuitive method for deciding whether to use AIC or likelihood cross-validation.

show abstract

Section: Introductionmentioning

confidence: 99%

A Simulation Study for the AIC and Likelihood Cross-validation: The Case of Exponential Versus Weibull Distributions

Takezawa

2018

JAMCS

View full text Add to dashboard Cite

show abstract

Asymptotic biases of information and cross-validation criteria under canonical parametrization

Cited by 1 publication

References 18 publications

A Simulation Study for the AIC and Likelihood Cross-validation: The Case of Exponential Versus Weibull Distributions

A Simulation Study for the AIC and Likelihood Cross-validation: The Case of Exponential Versus Weibull Distributions

Contact Info

Product

Resources

About