arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1706.00092cs.ITmath.IT
收藏

非精确梯度投影与快速数据驱动压缩感知

Inexact Gradient Projection and Fast Data Driven Compressed Sensing

Mohammad Golbabaee, Mike E. Davies

摘要:

我们研究了迭代投影梯度(IPG)算法对任意(可能非凸)集的收敛性,以及当梯度和投影预言都近似计算时的收敛性。我们考虑不同的概念,我们表明,渐进式固定精度(PFP)和$(1+ε)$最佳的预言可以达到相同的准确性,为准确的IPG算法的近似。我们表明,前计划也能够保持(线性)的精确算法的收敛速度,在相同的嵌入假设。相比之下,$(1+ε)$近似oracle需要更强的嵌入条件,适度的压缩比,它通常会减慢收敛速度。我们应用我们的结果来加速解决一类数据驱动的压缩感知问题,在那里我们取代迭代穷举搜索大数据集的快速近似最近邻搜索策略的基础上覆盖树数据结构。对于低内在维度的数据集,我们提出的算法实现了复杂性对数的数据集人口,而不是暴力搜索的线性复杂性。通过运行几个数值实验,我们得出了与理论分析预测相似的结论。

英文摘要:

We study convergence of the iterative projected gradient (IPG) algorithm for arbitrary (possibly nonconvex) sets and when both the gradient and projection oracles are computed approximately. We consider different notions of approximation of which we show that the Progressive Fixed Precision (PFP) and the $(1+ε)$-optimal oracles can achieve the same accuracy as for the exact IPG algorithm. We show that the former scheme is also able to maintain the (linear) rate of convergence of the exact algorithm, under the same embedding assumption. In contrast, the $(1+ε)$-approximate oracle requires a stronger embedding condition, moderate compression ratios and it typically slows down the convergence. We apply our results to accelerate solving a class of data driven compressed sensing problems, where we replace iterative exhaustive searches over large datasets by fast approximate nearest neighbour search strategies based on the cover tree data structure. For datasets with low intrinsic dimensions our proposed algorithm achieves a complexity logarithmic in terms of the dataset population as opposed to the linear complexity of a brute force search. By running several numerical experiments we conclude similar observations as predicted by our theoretical analysis.

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。

↑