Estimating Uncertainty of a Measurement Process

Abstract

Estimating a measurement of software quality is a challenge where uncertainty and variation have the greatest impact. Especially, at times when there is not enough information. Here, EUMP (estimating uncertainty of a measurement process) is introduced. EUMP is a recursive process which is using both multi regression and Monte Carlo simulation. It can be systematically obtained through EUMP for all distribution functions of both dependent and independent variables. Moreover, dependent variable can be estimated. Finally, a predictable system will be described, where the error of an estimated dependent variable will be proven it goes to zero when time (t) goes to infinity.