Characteristics of The Bayesian Information Criterion
- It is independent of the prior or the prior is "vague" (a constant).
- It can measure the efficiency of the parameterized model in terms of predicting the data.
- It penalizes the complexity of the model where complexity refers to the number of parameters in model.
- It is approximately equal to the minimum description length criterion but with negative sign.
- It can be used to choose the number of clusters according to the intrinsic complexity present in a particular dataset.
- It is closely related to other penalized likelihood criteria such as RIC and the Akaike information criterion.
Read more about this topic: Bayesian Information Criterion
Famous quotes containing the words characteristics of the, characteristics of, information and/or criterion:
“One of the characteristics of the dream is that nothing surprises us in it. With no regret, we agree to live in it with strangers, completely cut off from our habits and friends.”
—Jean Cocteau (1889–1963)
“Our day you will find that you have stopped regarding your baby as a totally unpredictable and therefore rather alarming novelty, and have begun instead to think of him as a person with tastes, preferences and characteristics of his own. When that happens you will know that he has moved on from being a “newborn” and has got himself settled into life.”
—Penelope Leach (20th century)
“I have all my life been on my guard against the information conveyed by the sense of hearing—it being one of my earliest observations, the universal inclination of humankind is to be led by the ears, and I am sometimes apt to imagine that they are given to men as they are to pitchers, purposely that they may be carried about by them.”
—Mary Wortley, Lady Montagu (1689–1762)
“I divide all literary works into two categories: Those I like and those I don’t like. No other criterion exists for me.”
—Anton Pavlovich Chekhov (1860–1904)