A Survey:A Multiple Comparisons Algorithm based Ranking and Clustering of COCOMO and Putnam's Software Cost Estimation Models

Software project can be completely predicting the most realistic effort using Software Cost Estimation. There are variety of methods and models trying to improve the estimation procedure of Software project development and application. From the variety of methods emerged the need for comparisons to determine the best model. Here, we propose a statistical framework based on a multiple comparisons algorithm in order to rank several cost estimation models, identifying those which have significant differences in accuracy, and clustering them in non overlapping groups. The proposed framework is applied in a large scale setup of comparing prediction models over datasets.