Sparse NonGaussian Component Analysis

Elmar Diederichs (1) , Anatoli B. Juditsky (2) , Vladimir Spokoiny (3) , Christof Schuette (1)

1 IMI - Institute for Mathematics and Informatics
2 SMS - Statistique et Modélisation Stochatisque
3 Weierstrass Institute and Humboldt University

Non-gaussian component analysis (NGCA) introduced in offered a method for high dimensional data analysis allowing for identifying a low-dimensional non-Gaussian component of the whole distribution in an iterative and structure adaptive way. An important step of the NGCA procedure is identification of the non-Gaussian subspace using Principle Component Analysis (PCA) method. This article proposes a new approach to NGCA called sparse NGCA which replaces the PCA-based procedure with a new the algorithm we refer to as convex projection.

Domaines: Statistiques [math.ST]; Théorie [stat.TH]

Elmar Diederichs, Anatoli B. Juditsky, Vladimir Spokoiny, Christof Schuette. Sparse NonGaussian Component Analysis. IEEE Transactions on Information Theory, 2010, 56 (6), pp.3033-3047. ⟨10.1109/TIT.2010.2046229⟩. ⟨hal-00381120⟩