Interval Type-2 Fuzzy Logic for Software Cost Estimation Using TSFC with Mean and Standard Deviation

Abstract

One of the biggest challenges in Software Engineering is accurately forecasting how much time and effort it will take either to develop a system. So far no model has proved to be successful at effectively and consistently predicting software development cost due to the lot of uncertainty factor of input size. In this paper we proposed an Interval Type 2 Fuzzy logic for software cost estimation. The inputs are fuzzified by using Takagi-Sugeno fuzzy controller of Universe Discourse with mean and standard deviation of size values affects the control performance. The software size can be regarded as a fuzzy set yielding the cost estimate also inform of a fuzzy set. The uncertainty is an inherit part in cost estimation. We reduce the uncertainty produced by the Type-1 functions by using Type-2 Fuzzy logic. We considered means FOU`s as a firing strength. The model deals effectively with imprecise and uncertain input and enhances the reliability of software cost estimation. The estimated effort is optimized using the developed model and tested on NASA software projects on the basis of three criterions for assessment of software cost estimation models. Comparison of the all models is done and it is found that the developed model provide better estimation.