Efficient Globally Convergent Stochastic Optimization for Canonical Correlation Analysis
arXiv:1604.01870
Abstract
We study the stochastic optimization of canonical correlation analysis (CCA), whose objective is nonconvex and does not decouple over training samples. Although several stochastic gradient based optimization algorithms have been recently proposed to solve this problem, no global convergence guarantee was provided by any of them. Inspired by the alternating least squares/power iterations formulation of CCA, and the shift-and-invert preconditioning method for PCA, we propose two globally convergent meta-algorithms for CCA, both of which transform the original problem into sequences of least squares problems that need only be solved approximately. We instantiate the meta-algorithms with state-of-the-art SGD methods and obtain time complexities that significantly improve upon that of previous work. Experimental results demonstrate their superior performance.
Accepted by NIPS 2016
References in corpus (9)
- A Universal Catalyst for First-Order Optimization
- Fast and Simple PCA via Convex Optimization
- Large-Scale Approximate Kernel Canonical Correlation Analysis
- Finding Linear Structure in Large Datasets with Scalable Canonical Correlation Analysis
- Large scale canonical correlation analysis with iterative least squares
- Scale Up Nonlinear Component Analysis with Doubly Stochastic Gradients
- Adaptive Canonical Correlation Analysis Based On Matrix Manifolds
- Robust Shift-and-Invert Preconditioning: Faster and More Sample Efficient Algorithms for Eigenvector Computation
- Principal Component Projection Without Principal Component Analysis