Artificial neural network for software reliability assessment

Neural networks have emerged as a promising technology in applications that require generalization, abstraction, adaptation and learning. The value of neural network modeling techniques in performing complicated pattern recognition and nonlinear forecasting tasks has now been demonstrated across an impressive spectrum of applications. This methodology has become an alternative to modelling some physical and non-physical systems with scientific or mathematical bases, and also an alternative to expert systems methodology. This paper presents an artificial neural network for software reliability assessment. It also discusses the main findings and concludes with promising results.