Parameterschätzung bei klassierten Daten

This paper considers the problem of parameter estimation when data of a random sample are given in the form of a frequency table. We give special consideration to a method that linearizes the cumulative distribution function. In that case parameters can be derived from the weighted estimation of a linear regression equation. The favourable properties of this estimation technique are demonstrated in a simulation experiment, where the parameters of a two-parameter-Weibull distribution are estimated