Input Data Analysis Using Neural Networks

Abstract

Simulation deals with real-life phenomena by constructing representative models of a system being questioned. Input data provide a driving force for such models. The requirement for iden tifying the underlying distributions of data sets is encountered in many fields and simulation applications (e.g., manufacturing economics, etc.). Most of the time, after the collection of the raw data, the true statistical distribution is sought by the aid of nonparametric statistical methods. In this paper, we investigate the feasi bility of using neural networks in selecting ap propriate probability distributions. The perfor mance of the proposed approach is measured with a number of test problems.