|
Abstract:
Since their introduction,
soft computing methods have found very wide range of practical applications,
often displaying superior performance in comparison to the traditional methods.
However, the application of soft computing tools has not been uniform, and it
did not reach yet some domains where traditional methods still prevail despite
their ineffectiveness. In this study we demonstrate advantage of utilizing a
soft computing tool: “Soft Regression” for designing effective demographic
policy. We conducted extensive literature survey and did not find any case of
soft computing applications utilized to design demographic policy; therefore we
consider this study as an initial introduction of soft computing to demographic
research. Soft
Regression (SR), is a modeling tool based on Soft Computing methods: Fuzzy
information processing and Heuristic approach. In contrast to traditional
statistical regression methods, it does not require restrictive conditions
(which often contradict the “real world” conditions), and thus avoids
computational distortions when such conditions are violated. It allows us to
include in the model all the relevant explanatory variables without losing some
variables due to multi-collinearity problem. Moreover, SR method performs
reliable computation of relative importance of the explanatory variables and
hence constitutes an effective tool for policy-makers to determine policy
priorities. There are additional advantages that will be explained later in the
article.
|
