Yashi Hu, Yue Zhang, Longteng Zhang, Xiangtian Zheng, Peichao Duan, Hard threshold acceleration algorithms for sparse signal recovery, Vol. 2024 (2024), Article ID 16, pp. 1-15

Full Text: PDF
DOI: 10.23952/jnfa.2024.16

Received Apri 24, 2024; Accepted July 22, 2024; Published August 22, 2024

 

Abstract. In sparse signal recovery, hard thresholding algorithms play an important role. By combining hard thresholding techniques with inertial acceleration, the heavy-ball-based hard thresholding algorithm (AIHT) and the heavy-ball-based hard thresholding pursuit algorithm (AHTP) have been proposed in recent years. On this basis, we propose two-step inertial acceleration algorithms A^2IHT and A^2HTP for signal recovery. In simulation experiments, we test the performance of A^2IHT and A^2HTP in signal recovery and image restoration as examples. The experimental results show that they take less time to recover with the two-step inertial acceleration algorithms, and the recovery effect is improved in some sense.

 

How to Cite this Article:
Y. Hu, Y. Zhang, L. Zhang, X. Zheng, P. Duan, Hard threshold acceleration algorithms for sparse signal recovery, J. Nonlinear Funct. Anal. 2024 (2024) 16.