Title :
Randomized recovery for boolean compressed sensing
Author :
Fatemi, Mehdi ; Vetterli, Martin
Author_Institution :
Lab. of Audiovisual Commun., Ecole Polytech. Fed. de Lausanne (EPFL), Lausanne, Switzerland
Abstract :
We consider the problem of boolean compressed sensing, which is also known as group testing. The goal is to recover a small number of defective items in a large set from a few collective binary tests. This problem can be formulated as a binary linear program, which is NP hard in general. To overcome the computational burden, it was recently proposed to relax the binary constraint on the variables, and apply a rounding to the solution of the relaxed linear program. In this paper, we introduce a randomized algorithm to replace the rounding procedure. We show that the proposed algorithm considerably improves the success rate with only a slight increase in computational cost.
Keywords :
compressed sensing; computational complexity; linear programming; Boolean compressed sensing; NP hard problem; binary linear programming; collective binary tests; randomized recovery algorithm; relaxed linear programming; Compressed sensing; Equations; Information theory; Mathematical model; Noise measurement; Testing; Vectors; Boolean compressed sensing; Group testing; Linear programming; Randomized algorithm;
Conference_Titel :
Information Theory Proceedings (ISIT), 2013 IEEE International Symposium on
Conference_Location :
Istanbul
DOI :
10.1109/ISIT.2013.6620270