Deep Learning for Augmented Visual Analytics: A Definitive Survey
Deep learning methods have significantly improved tasks across diverse domains ranging from machine translation to computer vision to self driving to computational biology. Building on this explosive success, researchers and practitioners in data visualization and visual analytics have started applying deep learning to automate various tasks and to intelligently augment user experience along the visual data analysis pipeline. In this paper, we give an overview of these promising but nascent efforts for augmented visual analytics using deep learning. We organize existing research and practice around tasks addressed and methods used, providing a structured view of the state of the art. We also discuss challenges informed by the limitations of the current approaches and outline opportunities and research directions to address these challenges.