Selection of the Best Statistical Distribution with Multicriteria Decision Making Technique TOPSIS

Abstract

Due to highly competitive global market, the organizations are now forced to focus more on increasing the productivity in their organization. Reliability estimation is necessary to meet that goal. Again for determining Reliability the Distributions of failure datasets are important. But which distribution will best represent that dataset is very important for the decision maker. In this paper an approach is adopted where the decision maker is the central actor and the method of selecting the best fit distributions is  by the multicriteria decision making tool TOPSIS. It is found that preference parameter lead to different Rankings. The preference parameter adopted are (D_MAX (Kolmogorov–Srimov test) deviation from empherical probability distribution function and deviation from empherical cumulative distribution function, deviation from skewness and the judgment by a group of specialist in the field of statistics). It has been found that exponential distribution is the best distribution to represent that dataset

Keywords: Best distribution, Multi-criteria decision-making, TOPSIS, [test for order preference by similarity to ideal solution], Decision making