É a questão da curiosidade, ou seja, você não está satisfeito com a minha resposta aqui ? Se não...
A investigação adicional dessa questão complicada mostrou que existe uma regra , que afirma que dois modelos são indistinguíveis pelo critério da se a diferença . O mesmo que você realmente lerá no artigo da wikipedia sobre (observe que o link é clicável!). Apenas para quem não clica nos links:AIC|AIC1−AIC2|<2AIC
AIC estima o apoio relativo a um modelo. Para aplicar isso na prática, começamos com um conjunto de modelos candidatos e, em seguida, localizamos os valores correspondentes dos modelos . Em seguida, identifique o valor mínimo da . A seleção de um modelo pode ser feita da seguinte maneira.AICAIC
Como regra geral, os modelos com suasAIC dentro de1–2 of the minimum have substantial support and should receive consideration in making inferences. Models having their AIC within about 4–7 of the minimum have considerably less support, while models with their AIC>10 above the minimum have either essentially no support and might be omitted from further consideration or at least fail to explain some substantial structural variation in the data.
A more general approach is as follows...
Denote the AIC values of the candidate models by AIC1, AIC2,AIC3,…,AICR. Let AICmin denotes the minimum of those values. Then e(AICmin−AICi)/2 can be interpreted as the relative probability that the i-th model minimizes the (expected estimated) information loss.
As an example, suppose that there were three models in the candidate set, with AIC values 100, 102, and 110. Then the second model is e(100−102)/2=0.368 times as probable as the first model to minimize the information loss, and the third model is e(100−110)/2=0.007 times as probable as the first model to minimize the information loss. In this case, we might omit the third model from further consideration and take a weighted average of the first two models, with weights 1 and 0.368, respectively. Statistical inference would then be based on the weighted multimodel.
Nice explanation and useful suggestions, in my opinion. Just don't be afraid of reading what is clickable!
In addition, note once more, AIC is less preferable for large-scale data sets. In addition to BIC you may find useful to apply bias-corrected version of AIC criterion AICc (you may use this R
code or use the formula AICc=AIC+2p(p+1)n−p−1, where p is the number of estimated parameters). Rule-of-thumb will be the same though.