Aufsatz(elektronisch)#1September 1991
Nonparametric density estimation with grouped observations
In: Statistica Neerlandica, Band 45, Heft 3, S. 255-269
ISSN: 1467-9574
A smooth and detailed distribution is fitted to coarsely grouped frequency data by a nonparametric approach, based on penalized maximum likelihood. The estimated distribution conserves mean and variance of the data. The numerical solution is described and a compact and simplified algorithm is given. The procedure is applied to two empirical datasets.