Link To Logit Models (logistic Regression)
Multinomial logit models, and certain other types of logistic regression, can be phrased as latent variable models with error variables distributed as Gumbel distributions (type I generalized extreme value distributions). This phrasing is common in the theory of discrete choice models, which include logit models, probit models, and various extensions of them, and derives from the fact that the difference of two type-I GEV-distributed variables follows a logistic distribution, of which the logit function is the quantile function. The type-I GEV distribution thus plays the same role in these logit models as the normal distribution does in the corresponding probit models.
Read more about this topic: Generalized Extreme Value Distribution
Famous quotes containing the words link and/or models:
“All successful men have agreed in one thing,—they were causationists. They believed that things went not by luck, but by law; that there was not a weak or a cracked link in the chain that joins the first and last of things.”
—Ralph Waldo Emerson (1803–1882)
“The parents who wish to lead a quiet life I would say: Tell your children that they are very naughty—much naughtier than most children; point to the young people of some acquaintances as models of perfection, and impress your own children with a deep sense of their own inferiority. You carry so many more guns than they do that they cannot fight you. This is called moral influence and it will enable you to bounce them as much as you please.”
—Samuel Butler (1835–1902)