First-Order Preconditioning via Hypergradient Descent
arXiv:1910.08461
Abstract
Standard gradient descent methods are susceptible to a range of issues that can impede training, such as high correlations and different scaling in parameter space.These difficulties can be addressed by second-order approaches that apply a pre-conditioning matrix to the gradient to improve convergence. Unfortunately, such algorithms typically struggle to scale to high-dimensional problems, in part because the calculation of specific preconditioners such as the inverse Hessian or Fisher information matrix is highly expensive. We introduce first-order preconditioning (FOP), a fast, scalable approach that generalizes previous work on hypergradient descent (Almeida et al., 1998; Maclaurin et al., 2015; Baydin et al.,2017) to learn a preconditioning matrix that only makes use of first-order information. Experiments show that FOP is able to improve the performance of standard deep learning optimizers on visual classification and reinforcement learning tasks with minimal computational overhead. We also investigate the properties of the learned preconditioning matrices and perform a preliminary theoretical analysis of the algorithm.
References in corpus (10)
- Adam: A Method for Stochastic Optimization
- Deep Residual Learning for Image Recognition
- Striving for Simplicity: The All Convolutional Net
- Optimizing Neural Networks with Kronecker-factored Approximate Curvature
- Gradient-based Hyperparameter Optimization through Reversible Learning
- Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions
- Preconditioned Stochastic Gradient Descent
- Natural Neural Networks
- Recurrent Deterministic Policy Gradient Method for Bipedal Locomotion on Rough Terrain Challenge
- A Kronecker-factored approximate Fisher matrix for convolution layers