Information Theoretic Limits for Linear Prediction with Graph-Structured Sparsity
Preprint
Abstract
We analyze the necessary number of samples for sparse vector recovery in a noisy linear prediction setup. This model includes problems such as linear regression and classification. We focus on structured graph models. In particular, we prove that sufficient number of samples for the weighted graph model proposed by Hegde and others is also necessary. We use the Fano's inequality on well constructed ensembles as our main tool in establishing information theoretic lower bounds.Keywords
All Related Versions
This publication has 4 references indexed in Scilit:
- Sample complexity for 1-bit compressed sensing and sparse classification2010
- Iterative hard thresholding for compressed sensingApplied and Computational Harmonic Analysis, 2009
- CoSaMP: Iterative signal recovery from incomplete and inaccurate samplesApplied and Computational Harmonic Analysis, 2009
- A Simple Proof of the Restricted Isometry Property for Random MatricesConstructive Approximation, 2008