English

Heavy-tailed Independent Component Analysis

Machine Learning 2015-09-03 v1 Statistics Theory Computation Machine Learning Statistics Theory

Abstract

Independent component analysis (ICA) is the problem of efficiently recovering a matrix ARn×nA \in \mathbb{R}^{n\times n} from i.i.d. observations of X=ASX=AS where SRnS \in \mathbb{R}^n is a random vector with mutually independent coordinates. This problem has been intensively studied, but all existing efficient algorithms with provable guarantees require that the coordinates SiS_i have finite fourth moments. We consider the heavy-tailed ICA problem where we do not make this assumption, about the second moment. This problem also has received considerable attention in the applied literature. In the present work, we first give a provably efficient algorithm that works under the assumption that for constant γ>0\gamma > 0, each SiS_i has finite (1+γ)(1+\gamma)-moment, thus substantially weakening the moment requirement condition for the ICA problem to be solvable. We then give an algorithm that works under the assumption that matrix AA has orthogonal columns but requires no moment assumptions. Our techniques draw ideas from convex geometry and exploit standard properties of the multivariate spherical Gaussian distribution in a novel way.

Keywords

Cite

@article{arxiv.1509.00727,
  title  = {Heavy-tailed Independent Component Analysis},
  author = {Joseph Anderson and Navin Goyal and Anupama Nandi and Luis Rademacher},
  journal= {arXiv preprint arXiv:1509.00727},
  year   = {2015}
}

Comments

30 pages

R2 v1 2026-06-22T10:47:33.254Z