We study independent component analysis with noisy observations.
We present, for the first time in the literature, consistent, polynomial-time algorithms to recover non-Gaussian source signals and the mixing matrix with
a reconstruction error that vanishes at a rate of T^1/2 using T observations and scales only polynomially with
the natural parameters of the problem.
Our algorithms and analysis also extend to deterministic source signals whose empirical distributions are approximately independent.
%0 Conference Paper
%1 HuGySze15
%A Huang, R.
%A György, A.
%A Szepesvári, Cs.
%B ICML
%D 2015
%K analysis complexity, component independent learning sample theory,
%P 2521--2530
%T Deterministic Independent Component Analysis
%X We study independent component analysis with noisy observations.
We present, for the first time in the literature, consistent, polynomial-time algorithms to recover non-Gaussian source signals and the mixing matrix with
a reconstruction error that vanishes at a rate of T^1/2 using T observations and scales only polynomially with
the natural parameters of the problem.
Our algorithms and analysis also extend to deterministic source signals whose empirical distributions are approximately independent.
@inproceedings{HuGySze15,
abstract = {We study independent component analysis with noisy observations.
We present, for the first time in the literature, consistent, polynomial-time algorithms to recover non-Gaussian source signals and the mixing matrix with
a reconstruction error that vanishes at a rate of T^{1/2} using T observations and scales only polynomially with
the natural parameters of the problem.
Our algorithms and analysis also extend to deterministic source signals whose empirical distributions are approximately independent.
},
added-at = {2020-03-17T03:03:01.000+0100},
author = {Huang, R. and Gy{\"o}rgy, A. and Szepesv{\'a}ri, {Cs}.},
biburl = {https://www.bibsonomy.org/bibtex/2e68fcdb51d00ccc2d484c68b4c6cb2dd/csaba},
booktitle = {ICML},
date-added = {2015-04-25 18:24:02 +0000},
date-modified = {2015-08-02 00:44:57 +0000},
interhash = {c7643defc251e88a29ca0c29617c935f},
intrahash = {e68fcdb51d00ccc2d484c68b4c6cb2dd},
keywords = {analysis complexity, component independent learning sample theory,},
pages = {2521--2530},
pdf = {papers/ICML15-DICA.pdf},
timestamp = {2020-03-17T03:03:01.000+0100},
title = {Deterministic Independent Component Analysis},
year = 2015
}