Tuning of COCO MO Model Parameter by using BEE Colony Optimization by Sherry Chalotra, cse
Abstract
Software estimation has become most necessary part of development process that is carried out at initial phases so as to get a rough idea of effort. Once effort is predicted it become easy to make predictions related to cost and time. But due to less availability, changing requirements, and incomplete data make it difficult to make such predictions. Many models are available but no one is capable of giving accurate results. The accurate results if obtained at early stages give significant benefits in budgeting, staffing management and control of project. This research work focused on effort estimation process and presented a method for estimating effort by using Bee Colony Optimization which tends to optimize COCOMO parameters. To measure the accuracy of proposed model MMRE is used and compared with COCOMO model and other existing models like, COCOMO II, SEL model, Halstead Model, Walston- Felix Model, and Bailey- Basil model. MMRE for COCOMO models comes out to be 0.43 whereas proposed model give values of 0.11 which showed that proposed model gave better results than COCOMO. When same proposed model was compared with other models the MMRE for it was optimum among all which suggests that proposed method can be used for planning of project, budgeting, and scheduling etc.
No comments:
Post a Comment