Real-Time Fault-Tolerant Computing with Machine Learning Enhancements

SUMMARY & CONCLUSIONSMachine learning enhancements applied to fault-tolerant computing are presented here. Classification of non-identical inputs and real-time environmental threats to fault-tolerant systems are handled using machine learning techniques. Efficient learning methods are developed so that they are suitable for real-time applications. The advantages of the enhancements are demonstrated through case studies. This work shows that not only the performance of machine learning can be improved by fault tolerance, but also fault-tolerant computing can benefit from machine learning.