Understanding AIC and BIC in Model Selection

By K. Burnham, David R. Anderson. 2004

4,849 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/ac05e354cebed9b37d350bac38f68a316fc35e72

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Functional diversity supports the physiological tolerance hypothesis for plant species richness along climatic gradients (2014, methodology): “As several quadratic relationships were found by Grace, Harrison and Damschen (2011), we tested for both linear and nonlinear (quadratic) relationships for each response measure and selected the best fit using Akaike Information Criteria (Burnham & Anderson 2004).”
Non‐stationary frequency analysis of extreme daily rainfall in Sao Paulo, Brazil (2009, methodology): “For a given model, the rAICc suggested by Burnham and Anderson (2004) and used in this work, for a quick and easy comparison and ranking of candidate models is rAICc = AICc − minAICc (12) where minAICc is the smallest AICc, regarding all the candidate models.”
Inferring Biological Networks by Sparse Identification of Nonlinear Dynamics (2016, background): “criteria would provide a systematic strength-of-evidence comparison between models [82].”
Linear regression (2012, unknown): “btT−1b ≤ c0 ⇐⇒ sup a =0 {( atb )2 atTa } ≤ c0, (16)”
Developing multiple hypotheses in behavioral ecology (2010, methodology): “…using Akaike’s information criterion (AIC), in which its wide use in ecology and evolutionary biology is largely due to influential works by David Anderson and Kenneth Burnham (Anderson et al. 2000, 2001; Burnham and Anderson 2001; Anderson and Burnham 2002; Burnham and Anderson 2002, 2004).”
Very High Resolution Spaceborne SAR Tomography in Urban Environment (2010, unknown): “The detailed derivation and performance of BIC are described in [32].”
Next‐generation monitoring of aquatic biodiversity using environmental DNA metabarcoding (2016, methodology): “The relative performance of these five models was compared using Akaike information criteria (Burnham & Anderson 2004).”
Rapid determination of comparative drought tolerance traits: using an osmometer to predict turgor loss point (2012, methodology): “Models were compared using the Akaike information criterion corrected for low n (AICc); the model with the lowest AICc value has best support, and differences >2 in AICc val- ues are considered meaningful (Burnham & Anderson 2002, 2004).”
The performance of African protected areas for lions and their prey (2017, unknown): “We started with univariate models of all covariates, and retained all models with some empirical support (ΔAICc of ≤7) (Burnham and Anderson, 2004).”
Identifying appropriate spatial scales of predictors in species distribution models with the random forest algorithm (2013, background): “The best fit is typically measured by the highest correlation coefficient or lowest AIC (Akaike information criterion) (Burnham & Anderson 2004).”
Sampling roots to capture plant and soil functions (2017, methodology): “The models were ranked according to the corrected Akaike information criterion (AICc) (Burnham & Anderson 2004).”
Improving data analysis in herpetology: Using Akaike's information criterion (AIC) to assess the strength of biological hypotheses (2006, unknown): “Alternative approaches include Bayesian methods and model selection based on information theory (Ellison, 1996; Wade, 2000; Burnham and Anderson, 2002, 2004).”
Methods for estimating temperature sensitivity of soil organic matter based on incubation data: A comparative evaluation (2015, methodology): “In the 1P model, although the parameters were constrained perfectly, the AIC analysis demonstrated that it is the worst model.”
Model selection and psychological theory: A discussion of the differences between the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC) (2012, background): “K-L divergence is just one kind of divergence measure, but the relative merits of K-L divergence are argued forcefully in Burnham and Anderson (2003, 2005), and has a somewhat intuitive interpretation in information theory.”
Estimating and comparing thermal performance curves (2006, background): “Two interrelated approaches are information theory and Bayesian theory (Burnham and Anderson, 2004; Ellison, 2004; Johnson and Omland, 2004).”