Abstract: Effort estimation of project development was a very challenging problem. It is simple and valid in using algorithmic approaches in effort estimation but they are not so reliable in many cases. Based on information collected from history of projects, the process are continuously improved in the project management to eliminate difficulties in estimating effort of software. However, several methods are used to estimate the software development effort accurately. These non-algorithmic models ware build up on limited number of resources and their performance have not been well investigated. This study compares algorithmic models like Doty, Bailey, Halstead, COCOMO Iand COCOMO II with non-algorithmic approaches like particleswarm optimization k-means clustering algorithms, triangular fuzzy approach and adaptive neuro fuzzy. The results suggests that the non-algorithmic approaches works well with more accurate and reliable.
N. Shivakumar, N. Balaji and K. Ansnthskumsr, 2016. A Comparative Analysis of Algorithmic and Soft Computing Techniques in Estimating Software Effort. Asian Journal of Information Technology, 15: 1207-1212.