Bayesian Software Prediction Models. Volume II. Classical and Bayesian Inference for the Software Imperfect Debugging Model.

Open Access
Publisher Website
Google Scholar
Abstract
This report presents two methods for statistical inference of the parameters of the imperfect debugging model proposed by Goel and Okumoto. Using the method of maximum likelihood, the mle's, the likelihood contours and the confidence regions for N, p and lambda are obtained. A Bayesian approach is presented to obtain the Bayesian point estimates and the H.P.D. regions. Numerical examples based on simulated data are used for illustrative purposes. (Author)

This publication has 0 references indexed in Scilit: