Analysis of Software Effort Estimation Based on Story Point and Lines of Code using Machine Learning

Abstract

The key responsibility of people involved in software project management is to estimate the software effort. Effort prediction is perplexing as the development of software is fluctuating. Several models have been developed in past 3 decades for software development cost estimation. Several cost estimation techniques, algorithmic models, non-algorithmic models and machine learning methods, exist. Machine learning methods are used with algorithmic or non-algorithmic models to get better accuracy. Researchers in past worked on the effort and time estimation using with one type of development methodology. In this paper, a comparative study has been done for agile development and traditional development using the neural network (NN) and genetic algorithm (GA). The minimum error and maximum accuracy for estimated values of effort achieved using the machine learning methods. The dataset with the story point give best results followed by projects with lines of code.