Data Driven Discovery of emergent behavior of engineering systems using interpretable deep learning
Abstract
Over the past few years data driven methodologies have grown to be learning a lot of of complicated information. Right from information of a simple partial differential equation, to complicated behavior of engineering systems. However, these approaches have been often rejected due to non-reliability of a black box methodologies. In this paper, we would consider how this notion can be weakened by the approach of Guided GradCAM and its associated features.