arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1207.4783cs.DScs.CC
收藏

测试永久性神谕--再访

Testing Permanent Oracles -- Revisited

Sanjeev Arora, Arnab Bhattacharyya, Rajsekar Manokaran, Sushant Sachdeva

摘要:

假设我们得到一个预言,它声称近似大多数矩阵X的积和式,其中X是从高斯系综中选择的(矩阵元素是独立同分布的)。单变量复高斯)。我们能证明神谕满足这一主张吗?本文给出了该任务的多项式时间算法。预言测试问题之所以有趣,是因为Aaronson和Arkhipov最近的一篇论文表明,如果存在一个多项式时间算法来模拟量子力学中的玻色子-玻色子相互作用,那么在BPP^NP中存在一个近似预言(上述类型)。由于计算偶数0/1矩阵的积和式是#P-完全的,这似乎证明了量子力学中比Shor的因式分解算法更强大的计算能力。然而,不像因式分解,这是在NP,这是不清楚以前如何测试的正确性近似预言的永久,这是贡献的文件。这里克服的技术困难是,一元多项式自校正,这是类似的预言测试算法的基础永久在有限领域--和其发现导致了复杂性理论的革命--似乎并不推广到复杂的(甚至,真正的)数字。我们相信,这个测试将推动进一步的进展,了解永久的高斯矩阵。

英文摘要:

Suppose we are given an oracle that claims to approximate the permanent for most matrices X, where X is chosen from the Gaussian ensemble (the matrix entries are i.i.d. univariate complex Gaussians). Can we test that the oracle satisfies this claim? This paper gives a polynomial-time algorithm for the task. The oracle-testing problem is of interest because a recent paper of Aaronson and Arkhipov showed that if there is a polynomial-time algorithm for simulating boson-boson interactions in quantum mechanics, then an approximation oracle for the permanent (of the type described above) exists in BPP^NP. Since computing the permanent of even 0/1 matrices is #P-complete, this seems to demonstrate more computational power in quantum mechanics than Shor's factoring algorithm does. However, unlike factoring, which is in NP, it was unclear previously how to test the correctness of an approximation oracle for the permanent, and this is the contribution of the paper. The technical difficulty overcome here is that univariate polynomial self-correction, which underlies similar oracle-testing algorithms for permanent over finite fields --- and whose discovery led to a revolution in complexity theory --- does not seem to generalize to complex (or even, real) numbers. We believe that this tester will motivate further progress on understanding the permanent of Gaussian matrices.

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。

↑