Estimating a mixing distribution in a multiple observation setting
Authors:
Y. P. Mack
Norman S. Matloff
Affiliation:
Division of Statistics, University of California, Davis, CA 95616, USA
Division of Computer Science, University of California, Davis, CA 95616, USA
Abstract:
Given unobservable i.i.d. r.v.'s Mi with c.d.f. F, and observable r.v.'s Yij with E(Yij|Mi) = Mi; we study a nonparametric estimate of F based on the Yij's by conditioning and kernel techniques. This investigation extends previous works of Blischke (1965) and James (1978).