On the Choice of a Model to Fit Data from an Exponential Family
Haughton, Dominique M. A.
Ann. Statist., Tome 16 (1988) no. 1, p. 342-355 / Harvested from Project Euclid
Let $X_1, \cdots, X_n$ be iid observations coming from an exponential family. The problem of interest is this: Given a finite number of models $m_j$ (smoothly curved manifolds in $\mathbb{R}^k$), choose the best model to fit the observations, with some penalty for choosing models with dimensions which are too large. A result of Schwarz is made more specific and is extended to the case where the models are curved manifolds. If $S(Y, n, j)$ is--up to a constant $C(n)$ independent of the model--the log of the posterior probability of the $j$th model, where the sample mean $Y_n = (1/n)\sum^n_{i=1}X_i$ has been replaced by $Y$, Schwarz suggested an asymptotic expansion of $S(Y, n, j)$ whose leading terms are $\gamma(Y, n, j) = n \sup_{\psi \in m_j \cap \Theta}(Y\psi - b(\psi)) - \frac{1}{2}k_j \log n$, in the case where the models are affine subspaces of $\mathbb{R}^k$. We establish a similar asymptotic expansion, including the next term, with uniform bounds for $Y$ in a compact neighborhood of $\nabla b(\theta)$, where $\theta$ is the true value of the parameter. We suggest a criterion for the choice of the best model that consists of maximizing the three leading terms in the expansion $S(Y, n, j)$. We show that the criterion gives the correct model with probabilities $P^n_\theta \rightarrow 1$ as $n \rightarrow + \infty$.
Publié le : 1988-03-14
Classification:  Curved models,  Schwarz's criterion,  exponential families,  62F12,  62H99
@article{1176350709,
     author = {Haughton, Dominique M. A.},
     title = {On the Choice of a Model to Fit Data from an Exponential Family},
     journal = {Ann. Statist.},
     volume = {16},
     number = {1},
     year = {1988},
     pages = { 342-355},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1176350709}
}
Haughton, Dominique M. A. On the Choice of a Model to Fit Data from an Exponential Family. Ann. Statist., Tome 16 (1988) no. 1, pp.  342-355. http://gdmltest.u-ga.fr/item/1176350709/