NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:5277
Title:Neural Networks with Cheap Differential Operators

The paper proposes deep neural networks architectures that are designed to allow for an efficient evaluation of differential operators. The approach is simple, original and likely to trigger further innovations in the coming years. It will also provide a baseline for future methods based on other approaches to solving the same type of problems.