Software Effort Prediction using Statistical and Machine Learning Methods
A. Ruchika Malhotra. International Journal of Advanced Computer Science and Applications(IJACSA), (2011)
Abstract
Accurate software effort estimation is an important part of software process. Effort is measured in terms of person months and duration. Both overestimation and underestimation of software effort may lead to risky consequences. Also, software project managers have to make estimates of how much a software development is going to cost. The dominant cost for any software is the cost of calculating effort. Thus, effort estimation is very crucial and there is always a need to improve its accuracy as much as possible. There are various effort estimation models, but it is difficult to determine which model gives more accurate estimation on which dataset. This paper empirically evaluates and compares the potential of Linear Regression, Artificial Neural Network, Decision Tree, Support Vector Machine and Bagging on software project dataset. The dataset is obtained from 499 projects. The results show that Mean Magnitude Relative error of decision tree method is only 17.06\%. Thus, the performance of decision tree method is better than all the other compared methods.
%0 Journal Article
%1 IJACSA.2011.020122
%A Ruchika Malhotra, Ankita Jain
%D 2011
%J International Journal of Advanced Computer Science and Applications(IJACSA)
%K decision effort estimation learning,oftware linear machine regression, tree,
%N 1
%T Software Effort Prediction using Statistical and Machine Learning Methods
%U http://ijacsa.thesai.org/
%V 2
%X Accurate software effort estimation is an important part of software process. Effort is measured in terms of person months and duration. Both overestimation and underestimation of software effort may lead to risky consequences. Also, software project managers have to make estimates of how much a software development is going to cost. The dominant cost for any software is the cost of calculating effort. Thus, effort estimation is very crucial and there is always a need to improve its accuracy as much as possible. There are various effort estimation models, but it is difficult to determine which model gives more accurate estimation on which dataset. This paper empirically evaluates and compares the potential of Linear Regression, Artificial Neural Network, Decision Tree, Support Vector Machine and Bagging on software project dataset. The dataset is obtained from 499 projects. The results show that Mean Magnitude Relative error of decision tree method is only 17.06\%. Thus, the performance of decision tree method is better than all the other compared methods.
@article{IJACSA.2011.020122,
abstract = { Accurate software effort estimation is an important part of software process. Effort is measured in terms of person months and duration. Both overestimation and underestimation of software effort may lead to risky consequences. Also, software project managers have to make estimates of how much a software development is going to cost. The dominant cost for any software is the cost of calculating effort. Thus, effort estimation is very crucial and there is always a need to improve its accuracy as much as possible. There are various effort estimation models, but it is difficult to determine which model gives more accurate estimation on which dataset. This paper empirically evaluates and compares the potential of Linear Regression, Artificial Neural Network, Decision Tree, Support Vector Machine and Bagging on software project dataset. The dataset is obtained from 499 projects. The results show that Mean Magnitude Relative error of decision tree method is only 17.06\%. Thus, the performance of decision tree method is better than all the other compared methods.},
added-at = {2014-02-21T08:00:08.000+0100},
author = {{Ruchika Malhotra}, Ankita Jain},
biburl = {https://www.bibsonomy.org/bibtex/2c6efca1ee4a9f7b929b783308c717cd9/thesaiorg},
interhash = {f05298336e751c2156d691c344c8fd2e},
intrahash = {c6efca1ee4a9f7b929b783308c717cd9},
journal = {International Journal of Advanced Computer Science and Applications(IJACSA)},
keywords = {decision effort estimation learning,oftware linear machine regression, tree,},
number = 1,
timestamp = {2014-02-21T08:00:08.000+0100},
title = {{Software Effort Prediction using Statistical and Machine Learning Methods}},
url = {http://ijacsa.thesai.org/},
volume = 2,
year = 2011
}