The Akaike information criterion is a measure of the relative goodness of fit of a statistical model. The AIC is grounded in the concept of information entropy, in effect offering a relative measure of the information lost when a given model is used to describe reality. It can be said to describe the tradeoff between bias and variance in model construction, or loosely speaking between accuracy and complexity of the model.
AIC values provide a means for model selection. AIC does not provide a test of a model in the sense of testing a null hypothesis; i.e. AIC can tell nothing about how well a model fits the data in an absolute sense. If all the candidate models fit poorly, AIC will not give any warning of that.
Read more about Akaike Information Criterion: Definition, How To Apply AIC in Practice, AICc, Relevance To Chi-squared Fitting, History, Bayesian Information Criterion
Famous quotes containing the words information and/or criterion:
“The family circle has widened. The worldpool of information fathered by the electric media—movies, Telstar, flight—far surpasses any possible influence mom and dad can now bring to bear. Character no longer is shaped by only two earnest, fumbling experts. Now all the world’s a sage.”
—Marshall McLuhan (1911–1980)
“There is only one art, whose sole criterion is the power, the authenticity, the revelatory insight, the courage and suggestiveness with which it seeks its truth.... Thus, from the standpoint of the work and its worth it is irrelevant to which political ideas the artist as a citizen claims allegiance, which ideas he would like to serve with his work or whether he holds any such ideas at all.”
—Václav Havel (b. 1936)