Estimating a difference of Kullback-Leibler risks using a normalized difference of AIC
From MaRDI portal
Abstract: AIC is commonly used for model selection but the precise value of AIC has no direct interpretation. We are interested in quantifying a difference of risks between two models. This may be useful for both an explanatory point of view or for prediction, where a simpler model may be preferred if it does nearly as well as a more complex model. The difference of risks can be interpreted by linking the risks with relative errors in the computation of probabilities and looking at the values obtained for simple models. A scale of values going from negligible to large is proposed. We propose a normalization of a difference of Akaike criteria for estimating the difference of expected Kullback-Leibler risks between maximum likelihood estimators of the distribution in two different models. The variability of this statistic can be estimated. Thus, an interval can be constructed which contains the true difference of expected Kullback-Leibler risks with a pre-specified probability. A simulation study shows that the method works and it is illustrated on two examples. The first is a study of the relationship between body-mass index and depression in elderly people. The second is the choice between models of HIV dynamics, where one model makes the distinction between activated CD4+ T lymphocytes and the other does not.
Recommendations
- Improvement to AIC as Estimator of Kullback–Leibler Information for Linear Model Selection
- Evaluation of the Kullback‐Leibler Discrepancy for Model Selection in Open Population Capture‐Recapture Models
- Information criteria and statistical modeling.
- Choice of estimators based on different observations: modified AIC and LCV criteria
- Choice of estimators based on Kullback-Leibler risk
Cites work
- Akaike's information criterion and recent developments in information complexity
- Bayesian model averaging: A tutorial. (with comments and a rejoinder).
- Bootstrap choice of estimators in parametric and semiparametric families: an extension of EIC
- Choice between Semi‐parametric Estimators of Markov and Non‐Markov Multi‐state Models from Coarsened Observations
- Estimating the dimension of a model
- Frequentist Model Average Estimators
- Generalised information criteria in model selection
- scientific article; zbMATH DE number 4088698 (Why is no real title available?)
- scientific article; zbMATH DE number 1034043 (Why is no real title available?)
- scientific article; zbMATH DE number 3444596 (Why is no real title available?)
- scientific article; zbMATH DE number 791392 (Why is no real title available?)
- scientific article; zbMATH DE number 3395163 (Why is no real title available?)
- Interaction in multidimensional contingency tables: an information theoretic approach
- Likelihood Ratio Tests for Model Selection and Non-Nested Hypotheses
- Mathematical and Computational Challenges in Population Biology and Ecosystems Science
- Maximum Likelihood Estimation in Dynamical Models of HIV
- Model selection and mixed-effects modeling of HIV infection dynamics
- Model Selection and Multimodel Inference
- MULTIPLE COMPARISONS OF LOG-LIKELIHOODS AND COMBINING NONNESTED MODELS WITH APPLICATIONS TO PHYLOGENETIC TREE SELECTION
- Optimal Model Assessment, Selection, and Combination
- Tests of Statistical Hypotheses Concerning Several Parameters When the Number of Observations is Large
Cited in
(22)- Choice of estimators based on different observations: modified AIC and LCV criteria
- Choice of prognostic estimators in joint models by estimating differences of expected conditional Kullback-Leibler risks
- Model selection using union-intersection principle for non nested models
- Joint modeling of the clinical progression and of the biomarkers' dynamics using a mechanistic model
- Estimation of the distribution of infection times using longitudinal serological markers of HIV: implications for the estimation of HIV incidence
- Empiricial comparison between some model selection criteria
- Tracking interval for selecting between non-nested models: an investigation for type II right censored data
- Treatment monitoring of HIV-infected patients based on mechanistic models
- Evaluation of the Kullback‐Leibler Discrepancy for Model Selection in Open Population Capture‐Recapture Models
- A distribution-free tracking interval for model selection: application in the strength of brittle materials
- Separated hypotheses testing for autoregressive models with non-negative residuals
- Parameter estimation and prediction of order statistics for the Burr type XII distribution with type II censoring
- Linear signed rank test for model selection
- Estimation and prediction for a unified hybrid-censored Burr type XII distribution
- Choice of estimators based on Kullback-Leibler risk
- Inference after separated hypotheses testing: an empirical investigation for linear models
- Modeling the dynamics of biomarkers during primary HIV infection taking into account the uncertainty of infection date
- Bootstrap choice of non-nested autoregressive model with non-normal innovations
- A universal approximate cross-validation criterion for regular risk functions
- Estimation of semi-Markov multi-state models: a comparison of the sojourn times and transition intensities approaches
- State selection in Markov models for panel data with application to psoriatic arthritis
- Statistical models: conventional, penalized and hierarchical likelihood
This page was built for publication: Estimating a difference of Kullback-Leibler risks using a normalized difference of AIC
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q958337)
