Computational Earth Sciences Group, Los Alamos National Laboratory, Los Alamos, NM, United States of America.
Department of Computer Science and Electrical Engineering, University of Maryland, Baltimore County, MD, United States of America.
PLoS One. 2021 Jan 6;16(1):e0244026. doi: 10.1371/journal.pone.0244026. eCollection 2021.
It was recently shown that quantum annealing can be used as an effective, fast subroutine in certain types of matrix factorization algorithms. The quantum annealing algorithm performed best for quick, approximate answers, but performance rapidly plateaued. In this paper, we utilize reverse annealing instead of forward annealing in the quantum annealing subroutine for nonnegative/binary matrix factorization problems. After an initial global search with forward annealing, reverse annealing performs a series of local searches that refine existing solutions. The combination of forward and reverse annealing significantly improves performance compared to forward annealing alone for all but the shortest run times.
最近的研究表明，量子退火可以作为某些类型矩阵分解算法中的有效、快速子程序。量子退火算法在快速、近似解方面表现最佳，但性能很快就达到了瓶颈。在本文中，我们在量子退火子例程中利用反向退火而不是正向退火来解决非负/二进制矩阵分解问题。在使用正向退火进行初始全局搜索之后，反向退火会执行一系列局部搜索，以改进现有解决方案。与仅使用正向退火相比，正向退火和反向退火的组合在所有情况下都显著提高了性能，除了最短的运行时间。
