Weight Optimization Based on Firefly Algorithm for Analogy-based Effort Estimation

Proper cost estimation is one of the vital tasks that must be achieved for software project development. Owing to complexity and uncertainties development process, this task ambiguous difficult. Recently, analogy-based (ABE) has become popular approaches in field due its effectiveness practicability comparing completed projects new estimating effort. However, spite many achievements, method not capable guarantee accurate confronting complex relation between independent features In such a case, performance ABE can improved by efficient feature weighting. This study introduces an enhanced integrating firefly algorithm (FA) with improving effort (SDEE). The proposed model provide identification similar optimising performances similarity function process which most relevant weights are assigned obtaining more estimates. A series experiments were carried out using six real-world datasets. results based on statistical analysis showed integration FA significantly outperformed existing especially ISBSG dataset.