An Effective Nature Inspired Approach for the Estimation of Software Development Cost

Accurate estimation of software cost and time is one of the important aspects of software project development. Accurate cost and effort predictions for software projects have a significant impact, meta-heuristic algorithms assist the software industry in predicting trustworthy and reliable values for software project planning and maintenance. Using these algorithms increases the software project success rate, in parallel with using the project development and management approaches of the current practices. Traditional software estimation techniques have deficiencies that are resolved by using meta-heuristic algorithms. Due to unclear model building and inaccurate results, few or none of these approaches are used. The motivation behind this research is to limit the gap between modern research results and executions inside organisations by proposing practical and effective nature-inspired algorithms for deployment and support approaches through the usage of research findings and industry-current practices. This research was accomplished by using the NASA 93 dataset, data selection, data distribution, and take-averaging on 3 meta-heuristic algorithms such as the Strawberry Algorithm (SBA), Gray Wolf Algorithm (GWO), and Harmony Search Algorithm (HSA). The inclusion criteria of these algorithms are high maturity, state-of-the-art, and representative. The purpose of algorithmic optimization is to minimise the Mean Magnitude of Relative Error (MMRE) and efficiently manage the activities of software development within estimated time and cost. These algorithms are evaluated based on their Magnitude of Relative Error (MRE) / Mean Magnitude of Relative Error (MMRE). GWO performs best in terms of MMRE reduction.