Abstract
DeRaS (Death Rates Simulations) is a new software application designed for the efficient life table construction under several detailed conditions defined by the user. User can choose from several methods of smoothing and extrapolation of the empirical mortality rates or probabilities of dying at higher ages.
In the first version of the application six well-known mortality models are included - the Gompertz and Gompertz-Makeham curve as two representatives of exponential functions, two logistic functions ("Kannisto" and "Thatcher" with constant) and two less frequently used models (Coale-Kisker, Heligman-Pollard) The unknown parameters are estimated in the nonlinear regression procedure where the weighted generalized least squares are used. The weights enable to eliminate the heteroskedasticity in data.