矩体中的最大熵
Maximal entropy in the moment body
AI总结:
研究确定点是否在矩体内的问题,基于量子信息论技术,提出全局最小化光滑严格凸对数划分函数的方法,通过数值实验揭示对矩阵大小的三次方依赖及预言机两种情况,指出主要瓶颈在于梯度存储和操作。
摘要:
矩体是谱复形(迹为1的半正定矩阵的凸集)的线性投影。确定给定点是否在给定矩体内在量子态估计和多项式优化中有诸多应用。矩体隶属预言机问题可用半定规划解决,有现成的内点求解器。本文受量子信息论技术启发,从分析和几何角度表明更有效的方法是全局最小化一个与最大熵问题对偶的光滑严格凸对数划分函数。分析了该函数的曲率性质,表明条件由点到矩体边界的距离决定,并描述了利用此分析的几何预处理算法。通过与内点和一阶半定求解器对比的基础数值实验,揭示了对矩阵大小的三次方依赖,类似于一些特征结构计算。还说明了预言机的两种情况:密集投影在几百大小内有效处理,稀疏实例如矩阵补全在标准笔记本上几分钟内可处理几千大小的矩阵。在这两种情况下,大规模半定规划方法的主要瓶颈几乎完全转移到高效梯度存储和操作上。
英文摘要:
A moment body is a linear projection of the spectraplex, the convex set of trace-one positive semidefinite matrices. Determining whether a given point lies within a given moment body is a problem with numerous applications in quantum state estimation and polynomial optimization. This moment body membership oracle can be addressed with semidefinite programming, for which several off-the-shelf interior-point solvers are available. In this paper, inspired by techniques from quantum information theory, we argue analytically and geometrically that a much more efficient approach consists of minimizing globally a smooth strictly convex log-partition function, dual to a maximum entropy problem. We analyze the curvature properties of this function, showing that conditioning is governed by the distance of the point to the boundary of the moment body, and we describe a neat geometric preconditioning algorithm that exploits this analysis. Basic numerical experiments, comparing against interior-point and first-order semidefinite solvers, reveal a cubic dependence on the matrix size, similar to a few eigenstructure computations. They also illustrate the two regimes of the oracle: dense projections are handled efficiently up to sizes of several hundred, while sparse instances such as matrix completion scale to matrices of size several thousand in minutes on a standard laptop. In both cases the main bottleneck in this approach to large-scale semidefinite programming is moved almost entirely to efficient gradient storage and manipulation.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。