ON THE ROLE OF UNBIASED INVERSE VARIANCE-STABILIZING TRANSFORMATION IN WAVELET SHRINKAGE ESTIMATION FOR NHPP-BASED SOFTWARE RELIABILITY ASSESSMENT
Xiao Xiao, Daichi Tada, Hisashi Yamamoto · International journal of industrial engineering · 2019
In non-homogeneous Poisson process (NHPP) based software reliability models (SRM), specifying the software intensity function or the mean value function accurately is important. A wavelet shrinkage estimation (WSE) method has been proposed for estimating NHPP-based SRMs in a non-parametric way. We concentrate on the inverse variance-stabilizing transformation (IVST) used in the last step of WSE and apply unbiased IVST instead of the commonly-used direct algebraic IVST to reduce the bias of the WSE estimator. Through numerical studies with real software fault count data, we show the effectiveness of the WSE combined with unbiased IVST.