In the 2018 NeurIPS conference, 4,845 papers were submitted. The paper I’m reviewing here by Chen et al, 2018, titled Neural Ordinary Differential Equations, won best paper award. The paper discusses using continuous Ordinary Differential Equations (ODE) for Neural Networks (NN) as opposed to the sorts of discrete layers used in the standard Recurrent Neural Networks (RNN).
Machine learning and deep learning has never been more advanced than it is today, with computing power evolving tremendously and becoming capable of processing information and learn from larger data sets than ever before. An emerging type of computing, known as neuromorphic computing, is making breakthroughs and becoming popular with AI and neuroscience researchers.