An efficient framework for practical Software Estimation with Machine Learning and Big Data Analytics
- B V A N S S Prabhakar Rao, JNTU, Kakinada, firstname.lastname@example.org
- P Seetha Ramaiah, College of Engineering (A), Andhra University, Visakhapatnam
Today, data analysis techniques in machine learning and statistics play a vital role not only in software industry but also in science & technology development. Software development effort is one of the most important metrics that must be correctly estimated in software projects. Almost no accessible model can estimate the cost of software with a high degree of efficiency. Every software industry has their own methodology for estimating their products
with experts. Estimating effort on the basis of expert judgment is the most common approach, and the decision to use such processes as an alternative of formal estimation models. This paper suggests a novel approach intended for better understanding on how to estimate the effort, cost and for all types of software projects, including fresh development, reengineering, and maintenance. Many industries have no separate processes and tools for estimating, planning, and bidding, but merging different processes to solve their problem. Reference to actual effort of similar projects if they exist with customer satisfaction to assess expected price-to-win is more useful for estimators. Software managers treat planning and bidding the project as processes separate from effort estimation.
In real scenario planning is purely depend on the estimation. They need to tie together, and then only we could reach better results. Machine Learning and Data Analytical techniques will be helpful in solving this kind of problem.