Boehm and Valerdi (2006) reported that the uncertainty level of effort estimation during
early stage of software development is very high. The uncertainty present in the attribute
causes some inaccuracy in the prediction [url removed, login to view] improve the performance of analogybased estimation at the early stage, combined at with fuzzy numbers and calculate the similarity distances. But sometimes the most similar project still has a similarity distance with the project being estimated in most of the cases. Essentially, the effort needs to be adjusted when the most similar project has a similarity distance with the project being estimated. To adjust the effort in Estimation by Analogy, the proposed approach uses Genetic Algorithm method.