Rates of convergence for the maximum likelihood estimator in mixture models
Sara van de Geer · Journal of nonparametric statistics · 1996
This paper studies the maximum likelihood estimator of an unknown density , where is a given convex class of densities. In a mixture model, can be seen as the convex hull of a collection of kernels. It is shown that the dimension of and the behaviour of f 0 near zero determine a rate of convergence in Hellinger distance of .