Abstract: The pridiction of future failure data from observed data sets can been transformed into a problem of nolinear regression, and the kernel functions method is very efficient for solving nolinear regression problems. A kernel func- lions-based generic model adaptive to the characteristic of the given data sets is used for software failure time predic- lion, and it is applied to learn and recognize the inherent internal temporal property of software failure sequence in order to capture the most current feature hidden inside the software failure behavior. The experimental results based on four- teen real data sets show that the proposed model has better prediction and applicability than that of some other condi- tional software reliability prediction models.

Key words: Kernel functions,Kernel regression,Software reliability model,Nonhomogeneous poisson process