AI researchers devise failure detection method for safety-critical machine learning

Researchers from MIT, Stanford University, and the University of Pennsylvania have devised a method for predicting failure rates of safety-critical machine learning systems and efficiently determining their rate of occurrence.

Read More