11th IEEE Symposium on Computers and Communications (ISCC'06)
Wavelet Based RDNN for Software Reliability Estimation
Cagliari, Sardinia, Italy
June 26-June 29
ISBN: 0-7695-2588-1
Using the wavelet basis in Recurrent Dynamic Neural Network (RDNN) can improve the failure event estimation of software defect tracking in telecommunications. Non-linearity of the system is represented by proper selection of the wavelet function. This RDNN handles noisy data and enhances the speed of convergence as compared with alternate approaches. A new adaptive RDNN is presented where software deployment testing observations are used to synthesize intrinsic model parameters.
Citation:
Adam Smiarowski, Jr., Hoda S. Abdel-Aty-Zohdy, Mostafa Hashem Sherif, Hemal Shah, "Wavelet Based RDNN for Software Reliability Estimation," iscc, pp.312-317, 11th IEEE Symposium on Computers and Communications (ISCC'06), 2006