Modeling Data with Flat Densities

In this article, a simple probability distribution for modeling data with flat densities is introduced and used to model real world assessment data. The new distribution is a mixture of three distributions, two truncated normals, and a uniform. The parameters are estimated by using the sample percentiles because the likelihood and the method of moment approaches result in complicated forms. The proposed method is mainly based on the shape of the empirical density. The proposed method looks promising for modeling flat densities that are similar to the one used in the study.