English

PCA with Gaussian perturbations

Machine Learning 2015-07-27 v2 Machine Learning

Abstract

Most of machine learning deals with vector parameters. Ideally we would like to take higher order information into account and make use of matrix or even tensor parameters. However the resulting algorithms are usually inefficient. Here we address on-line learning with matrix parameters. It is often easy to obtain online algorithm with good generalization performance if you eigendecompose the current parameter matrix in each trial (at a cost of O(n3)O(n^3) per trial). Ideally we want to avoid the decompositions and spend O(n2)O(n^2) per trial, i.e. linear time in the size of the matrix data. There is a core trade-off between the running time and the generalization performance, here measured by the regret of the on-line algorithm (total gain of the best off-line predictor minus the total gain of the on-line algorithm). We focus on the key matrix problem of rank kk Principal Component Analysis in Rn\mathbb{R}^n where knk \ll n. There are O(n3)O(n^3) algorithms that achieve the optimum regret but require eigendecompositions. We develop a simple algorithm that needs O(kn2)O(kn^2) per trial whose regret is off by a small factor of O(n1/4)O(n^{1/4}). The algorithm is based on the Follow the Perturbed Leader paradigm. It replaces full eigendecompositions at each trial by the problem finding kk principal components of the current covariance matrix that is perturbed by Gaussian noise.

Keywords

Cite

@article{arxiv.1506.04855,
  title  = {PCA with Gaussian perturbations},
  author = {Wojciech Kotłowski and Manfred K. Warmuth},
  journal= {arXiv preprint arXiv:1506.04855},
  year   = {2015}
}
R2 v1 2026-06-22T09:54:17.962Z