Optimization of Software Cost Estimation Model Based on Present Past Future Algorithm

Accurate Software Cost Estimation (SCE) is a significant challenge in Software Engineering, particularly due to the limitations of traditional models in handling complex modern projects. To overcome these challenges, there is a rising trend in utilizing nature-inspired meta-heuristic algorithms. Among these, the Constructive Cost Model (COCOMO) is a prominent regression-based method for estimating software costs. However, COCOMO's fixed coefficients, which do not adapt to project variations, pose a limitation. To address this, we introduce a novel approach using the Past Present Future (PPF) algorithm, which leverages historical project data to optimize cost estimation models. We assess the model's performance using standard evaluation metrics, such as the mean magnitude of relative error (MMRE) and the Manhattan distance (MD). Our research demonstrates the effectiveness of the PPF-based optimization in improving cost estimation accuracy compared to other methods. This contribution enhances the applicability of SCE in modern software development, offering practical insights for project managers and practitioners.