Given a mixture of binomial distributions, how do we estimate the
unknown mixing distribution? We build on earlier work of Lindsay and further
elucidate the geometry underlying this question, exploring the approximating
role played by cyclic polytopes. Convergence of a resulting maximum likelihood
fitting algorithm is proved and numerical examples given; problems over the
lack of identifiability of the mixing distribution in part disappear.
Publié le : 1999-10-14
Classification:
Binomial,
mixture,
mixing distribution,
geometry,
moment curve,
cyclic polytope,
nearest point,
least squares,
weighted least squares,
maximum likelihood,
Kullback-Leibler distance,
62G99,
62P15,
52B12
@article{1017939148,
author = {Wood, G. R.},
title = {Binomial mixtures: geometric estimation of the mixing
distribution},
journal = {Ann. Statist.},
volume = {27},
number = {4},
year = {1999},
pages = { 1706-1721},
language = {en},
url = {http://dml.mathdoc.fr/item/1017939148}
}
Wood, G. R. Binomial mixtures: geometric estimation of the mixing
distribution. Ann. Statist., Tome 27 (1999) no. 4, pp. 1706-1721. http://gdmltest.u-ga.fr/item/1017939148/