A Nonparametric Regression Software Reliability Growth Model Based on Two Estimators
Yii, Toh, Panlop Zeephongsekul · RMIT Research Repository (RMIT University Library) · 2009
In this paper, a new estimator combining two Nonparametric Regression (NPR) estimators, Nadaraya-Watson (NW) and Local Linear (LL) estimators, is used to predict software reliability growth. The fact that is does not impose any underlying parametric model on the data structure makes its application straightforward and simple as opposed to the parametric models used in most SRGMs. We also demonstrate that the new estimator outperforms some tradtitional parametric SRGMs in its prediction of software reliability growth.