Abstract: Studies have shown that deep learning models are vulnerable to adversarial examples, which cause incorrect predictions by adding imperceptible perturbations into normal inputs. The same ...

Abstract: The survival signature, a new approach introduced by Coolen, has been used as an effective method to quantify the complex system's reliability as it requires less computation cost than other ...