DMCA
An Empirical Process for Building and Validating Software Engineering Parametric Models (2005)
Cached
Download Links
| Citations: | 3 - 2 self |
BibTeX
@MISC{Sherriff05anempirical,
author = {Mark Sherriff and Barry Boehm and Laurie Williams and et al.},
title = {An Empirical Process for Building and Validating Software Engineering Parametric Models},
year = {2005}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Parametric modeling is a statistical technique whereby a dependent variable is estimated based on the values of and the relationships between the independent variable(s). The nature of the dependent variable can vary greatly based on one’s domain of interest. In software engineering, parametric models are often used to help predict a system’s development schedule, cost-to-build, and quality at various stages of the software lifecycle. In this paper, we discuss the use of parametric modeling in software engineering and present a nine-step parametric modeling process for creating, validating, and refining software engineering parametric models. We illustrate this process with three software engineering parametric models. Each of these models followed the nine-steps in different ways due to the research technique, the nature of the model, and the variability of the data. The three models have been shown to be effective estimators of their respective independent variables. This paper aims to assist other software engineers in creating parametric models by establishing important steps in the modeling process and by demonstrating three variations on following the nine-step process.
Keyphrases
validating software engineering parametric model empirical process software engineering parametric model software engineering parametric model dependent variable parametric modeling various stage different way nine-step process software lifecycle respective independent variable independent variable research technique statistical technique effective estimator nine-step parametric modeling process important step software engineer system development schedule modeling process
