Parameter Tuning of Software Effort Estimation Models Using Antlion Optimization
| Main Author: | Zedan, Marrwa |
|---|---|
| Format: | Article info eJournal |
| Bahasa: | eng |
| Terbitan: |
Universitas Ahmad Dahlan
, 2020
|
| Subjects: | |
| Online Access: |
http://journal.uad.ac.id/index.php/TELKOMNIKA/article/view/16907 |
Daftar Isi:
- In this work, four modified models are used to estimate the effort required by a project. These models are modifications of the well-known Constructive Cost Model (COCOMO). The Antlion Optimization (ALO) Algorithm is employed to estimate the parameters of the models due to its efficiency and wide applicability in many areas.Three tests are carried out to show the efficiency of ALO: first, ALO is used with Bailey and Basili dataset to optimize parameters for the basic COCOMO Model and Sheta’s Model l and 2, and was compared with the Firefly Algorithm (FA), Genetic Algorithms (GA), and Particle Swarm Optimization (PSO). Second, the parameters of Sheta’s Model 1 and 2, Uysal’s Model 1 and 2 are optimized using Bailey and Basili dataset; results were compared with Directed Artificial Bee Colony Algorithm (DABCA), GA, and Simulated Annealing (SA). The third test involves using ALO to estimate the parameters of the Basic COCOMO model using four large datasets, results are compared with hybrid Bat inspired Gravitational Search Algorithm method called (hBATGSA), the improved BAT (IBAT), and BAT algorithms. Results of Tests 1 and 2 show that ALO outperformed other algorithms, as for Test 3, ALO was better than BAT and IBAT in terms of MAE and the number of best estimated projects. ALO was able to achieve better results than hBATGSA for Dataset2 and Dataset4 out of the four datasets explored also in terms of MAE and the number of best estimated projects for each dataset.