Zare Baghiabad, Fatemeh, Khademi Zare, Hasan, Fallahnezhad, Mohammadsaber, Adibnia, Fazlollah. (1396). Estimation of total Effort and Effort Elapsed in Each Step of Software Development Using Optimal Bayesian Belief Network. , 9(3), 507-530. doi: 10.22059/jitm.2017.61641
Fatemeh Zare Baghiabad; Hasan Khademi Zare; Mohammadsaber Fallahnezhad; Fazlollah Adibnia. "Estimation of total Effort and Effort Elapsed in Each Step of Software Development Using Optimal Bayesian Belief Network". , 9, 3, 1396, 507-530. doi: 10.22059/jitm.2017.61641
Zare Baghiabad, Fatemeh, Khademi Zare, Hasan, Fallahnezhad, Mohammadsaber, Adibnia, Fazlollah. (1396). 'Estimation of total Effort and Effort Elapsed in Each Step of Software Development Using Optimal Bayesian Belief Network', , 9(3), pp. 507-530. doi: 10.22059/jitm.2017.61641
Zare Baghiabad, Fatemeh, Khademi Zare, Hasan, Fallahnezhad, Mohammadsaber, Adibnia, Fazlollah. Estimation of total Effort and Effort Elapsed in Each Step of Software Development Using Optimal Bayesian Belief Network. , 1396; 9(3): 507-530. doi: 10.22059/jitm.2017.61641

Estimation of total Effort and Effort Elapsed in Each Step of Software Development Using Optimal Bayesian Belief Network

Journal of Information Technology Management
مقاله 5، دوره 9، شماره 3، 2017، صفحه 507-530    اصل مقاله (472.62 K)
نوع مقاله: Research Paper
شناسه دیجیتال (DOI): 10.22059/jitm.2017.61641
نویسندگان
Fatemeh Zare Baghiabad1؛ Hasan Khademi Zare* 2؛ Mohammadsaber Fallahnezhad2؛ Fazlollah Adibnia3
1Ph.D. Candidate Industrial Engineering, Yazd University, Yazd, Iran
2Associate Prof., Dep. of Industrial Engineering, Yazd University, Yazd, Iran
3Assistant Prof., Dep. of Computer Engineering, Yazd University, Yazd, Iran
چکیده
Accuracy in estimating the needed effort for software development caused software effort estimation to be a challenging issue. Beside estimation of total effort, determining the effort elapsed in each software development step is very important because any mistakes in enterprise resource planning can lead to project failure. In this paper, a Bayesian belief network was proposed based on effective components and software development process. In this model, the feedback loops are considered between development steps provided that the return rates are different for each project. Different return rates help us determine the percentages of the elapsed effort in each software development step, distinctively. Moreover, the error measurement resulted from optimized effort estimation and the optimal coefficients to modify the model are sought. The results of the comparison between the proposed model and other models showed that the model has the capability to highly accurately estimate the total effort (with the marginal error of about 0.114) and to estimate the effort elapsed in each software development step.
کلیدواژه‌ها
Bayesian belief network؛ Enterprise resource planning؛ Machine learning methods؛ Software effort estimation؛ Optimization
مراجع
آمار
تعداد مشاهده مقاله: 1,436
تعداد دریافت فایل اصل مقاله: 1,141