HoD-Net: High-Order Differentiable Deep Neural Networks and Applications

We introduce a deep architecture named HoD-Net to enable high-order differentiability for deep learning. HoD-Net is based on and generalizes the complex-step finite difference (CSFD) method. While similar to classic finite difference, CSFD approaches the derivative of a function from a higher-dimension complex domain, leading to highly accurate and robust differentiation computation without numerical stability issues. This method can be coupled with backpropagation and adjoint perturbation methods for an efficient calculation of high-order derivatives. We show how this numerical scheme can be leveraged in challenging deep learning problems, such as high-order network training, deep learningbased physics simulation, and neural differential equations.

Authors

• Siyuan Shen
• Tianjia Shao State Key Lab of CAD&CG, Zhejiang University
• Kun Zhou State Key Lab of CAD&CG, Zhejiang University, Hangzhou, China
• Chenfanfu Jiang
• Feng Luo
• Yin Yang

Keywords

No keywords are indexed for this paper.