An Inverse Power Method for Nonlinear Eigenproblems with Applications in 1-Spectral Clustering and Sparse PCA

TL;DR

This paper introduces a generalized inverse power method for solving nonlinear eigenproblems, demonstrating its effectiveness in 1-spectral clustering and sparse PCA with state-of-the-art results.

Contribution

It presents a novel inverse power method for nonlinear eigenproblems, extending traditional eigenvalue techniques to broader applications in machine learning.

Findings

Achieves state-of-the-art results in 1-spectral clustering

Achieves state-of-the-art results in sparse PCA

Method guarantees convergence to a nonlinear eigenvector

Abstract

Many problems in machine learning and statistics can be formulated as (generalized) eigenproblems. In terms of the associated optimization problem, computing linear eigenvectors amounts to finding critical points of a quadratic function subject to quadratic constraints. In this paper we show that a certain class of constrained optimization problems with nonquadratic objective and constraints can be understood as nonlinear eigenproblems. We derive a generalization of the inverse power method which is guaranteed to converge to a nonlinear eigenvector. We apply the inverse power method to 1-spectral clustering and sparse PCA which can naturally be formulated as nonlinear eigenproblems. In both applications we achieve state-of-the-art results in terms of solution quality and runtime. Moving beyond the standard eigenproblem should be useful also in many other applications and our inverse power method can be easily adapted to new problems.