Paper Info
Title
Estimating software maintenance effort: a neural network approach
Abstract
Software maintenance forms an essential component of software development. Its planning includes estimation of maintenance effort, duration, personnel and costs. Adequate information regarding size, complexity and maintainability is however often unavailable. In the present work, a Neural Network (NN) based effort estimator is developed using Matlab. A feed forward back- propagation NN employing Bayesian regularization training is selected and trained for one dataset. Various categories of software maintenance cost drivers and their effect on maintenance effort have been analyzed using different combinations of number of hidden layers and hidden neurons etc. The NN is able to successfully model the maintenance effort as the obtained results are well within the previously published error limits
Year
DOI
Venue
2008
10.1145/1342211.1342232
ISEC
Keywords
Field
DocType
neural network approach,software maintenance,estimating software maintenance effort,adequate information,software maintenance cost driver,hidden layer,effort estimator,maintenance effort,neural network,software development,bayesian regularization training,propagation nn,feed forward,bayesian regularization,back propagation
Data mining,Analysis effort method,Computer science,Cost driver,Artificial intelligence,Software maintenance,Software metric,Artificial neural network,Maintainability,Software development,Machine learning,Software sizing
Conference
Citations 
PageRank 
References 
10
0.54
14
Authors
2
Name
Order
Citations
PageRank
Ruchi Shukla1152.32
Arun Kumar Misra2426.08