EA-CG: An Approximate Second-Order Method for Training Fully-Connected Neural Networks
- Resource Type
- Working Paper
- Authors
- Chen, Sheng-Wei; Chou, Chun-Nan; Chang, Edward Y.
- Source
- Subject
- Computer Science - Machine Learning
- Language
For training fully-connected neural networks (FCNNs), we propose a practical approximate second-order method including: 1) an approximation of the Hessian matrix and 2) a conjugate gradient (CG) based method. Our proposed approximate Hessian matrix is memory-efficient and can be applied to any FCNNs where the activation and criterion functions are twice differentiable. We devise a CG-based method incorporating one-rank approximation to derive Newton directions for training FCNNs, which significantly reduces both space and time complexity. This CG-based method can be employed to solve any linear equation where the coefficient matrix is Kronecker-factored, symmetric and positive definite. Empirical studies show the efficacy and efficiency of our proposed method.
Comment: Change to AAAI-19 Version