Mobile systems and devices including Android are vulnerable to the effects of software aging which are manifested in performance degradation during long run-time. It is important to identify efficient system and user interface metrics for detecting and counteracting the software aging effects. The aging metrics used in researches of the Android operating system do not take into account the aging processes in user applications.
In this paper the radial-basis neural network was used for software failures prediction. The influence of activation function of the RBF neural net on the learning efficiency and software failures prediction is studied. It is shown that the optimal activation function is Inverse Multiquadric with 10 neurons in the input layer and 30 neurons in the hidden one (square of Pearson correlation coefficient is 0.997 and mean deviation is 14.4).
Assumption of independent components execution in software reliability models built using architectural approach is a simplification of real software execution. In this paper Gokhale model with higher order Markov chains has been improved to appreciate software execution dependencies in it's reliability prediction.