## Conjugate Gradient Iterative Hard Thresholding

The development of this software and the performance analysis of NIHT, HTP, and CSMPSP has led to the development of a new algorithm which exploits the computational advantages of
these algorithms. The CGIHT family of algorithms is introduced in the following paper along with theoretical, uniform recovery guarantees and extensive empirical testing.

The performance of the algorithms under the model y=Ax+e for a misfit or noise vector e is provided in the following paper.
**CGIHT: Conjugate Gradient Iterative Hard Thresholding for compressed sensing and matrix completion**

J.D. Blanchard, J. Tanner, K. Wei

Information and Inference: a journal of the IMA, in press, 2015.J.D. Blanchard, J. Tanner, K. Wei

Information and Inference: a journal of the IMA, in press, 2015.

**Conjugate Gradient Iterative Hard Thresholding: observed noise stability in compressed sensing**

J.D. Blanchard, J. Tanner, K. Wei

IEEE Transactions on Signal Processing, 63(2): 528.537, 2015.

J.D. Blanchard, J. Tanner, K. Wei

IEEE Transactions on Signal Processing, 63(2): 528.537, 2015.

## Expander L-0 Decoding

**R. Mednozo-Smith, J. Tanner**

Submitted 2015.

Abstract:Two new algorithms, Serial L-0 and Parallel L-0, are introduced for use in the setting of combinatorial compressed sensing with expander matrices. These algorithms are shown to outperform existing algorithms with Parallel L-0 capable of solving solving extremely large problems in seconds.

Submitted 2015.

Abstract: