Imprecision and Uncertainty in Software Effort Prediction Models : FUZZY LOGIC BASED FRAMEWORK FOR HANDLING IMPRECISION AND UNCERTAINTY IN SOFTWARE DEVELOPMENT EFFORT PREDICTION MODELS

The prediction of software development effort from early software estimates is a challenging task because not much of the information about the software is available at that time. Moreover, the information gathered for software attributes from various sources is subjective to imprecision and uncertainty. Imprecision arises when an expert defines some qualitative criteria to differentiate between two or more classes. Uncertainty arises due to the existence of more than one metric for a particular attribute. This uncertainty leads to the uncertainty in software effort prediction. This book presents a type-2 Fuzzy Logic System (FLS) based effort prediction framework that is capable of handling imprecision and uncertainty. In addition, empirical studies regarding the impact of various parameters of FLS on effort prediction system are carried out. The book also identifies a set of attributes that can give indication of the credibility of existing size metrics; and uses these attributes to evaluate some existing software size metrics.