Estimating a distribution function with truncated data (Q5906533): Difference between revisions

The author considers a special truncation model for incomplete observations occurring for example in astronomy and derives a nonparametric maximum likelihood estimator for the unknown underlying distribution function. This estimator is a modification of an estimator proposed by \textit{D. A. Linden-Bell} [Mon. Nat. R. Astr. Soc. 155, 95-118 (1971)]. The strong pointwise and uniform consistency of this modified estimator is investigated and sufficient conditions for the validity of laws of the iterated logarithm are stated. Unfortunately the paper contains such typographical errors that it is difficult to find one of the main theorems (Theorem 3.1) and to put it in the correct form.