摘要：Cooperative neural networks have been used for the last few decades in a broad variety of applications. If the application is conclusive or decision based, like in classification or clustering problems, then it is suitable to use. While measuring the reliability or the failure aftermaths, it has been proved that essential constituent which assist in enhancing the reliability is having no unitary point of failure in a cooperative neural network. In this paper we have studied the reliability of any cooperative neural network and made a review of the possibility of implementing probabilistic mathematical model to check whether a cooperative neural network is reliable and few general results are provided for the reliability evaluation
