Software Testing Cost Estimation Based On Neural Network And Rough Sets

As software quality has got more attention and higher requirement than before, the development and testing of software have been separated in most cases now. More and more software companies choose to out-source their testing work to those professional companies. Software testing is an important part for software quality assurance.This paper first introduces the situation and characteristics of the software testing industry, and compares the different methods for software estimation. Then it has established a new model for cost estimation of software testing project based on neural network and rough sets theory. Finally, it gives an example to verify this model is applicable in practice, and analyze its advantages and disadvantages.