It is commonly assumed that neural networks have a built-in fault tolerance property mainly due to their parallel structure. The international community of neural networks discussed these properties until 1994 and afterward the subject has been mostly ignored. Recently, the subject was again brought to discussion due to the possibility of using neural networks in areas where fault tolerance and graceful degradation properties would be an added value, like medical applications of nano-electronics or space missions. Nevertheless, the evaluation of fault tolerance and graceful degradation characteristics remained difficult because there were no systematic methods or tools that could be easily applied to a given Artificial Neural Networks application. The discussion of models is the first step for sorting ways of developing the fault tolerance capability and for building a tool that can evaluate and improve this characteristic. The present work proposes a fault tolerance model, presents solutions for improving it and introduces the Fault Tolerance Simulation and Evaluation Tool for Artificial Neural Networks that evaluates and improves fault tolerance.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.