arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1206.6409cs.LGcs.DCstat.ML
收藏

求解大型$\ell_1$正则化问题的放大坐标下降算法

Scaling Up Coordinate Descent Algorithms for Large $\ell_1$ Regularization Problems

Pacific Northwest National Laboratory(太平洋西北国家实验室);University of Texas at Austin(德克萨斯大学奥斯汀分校)

Chad Scherrer, Mahantesh Halappanavar, Ambuj Tewari, David Haglin

摘要:

我们提出了一个通用的框架,并行坐标下降(CD)算法,其中包括,作为特殊情况下,原来的顺序算法循环CD和随机CD,以及最近的并行猎枪算法。我们介绍了两种新的并行算法,也是特殊情况-线程贪婪的CD和基于着色的CD-并给出了这些的OpenMP实现的性能测量。

英文摘要:

We present a generic framework for parallel coordinate descent (CD) algorithms that includes, as special cases, the original sequential algorithms Cyclic CD and Stochastic CD, as well as the recent parallel Shotgun algorithm. We introduce two novel parallel algorithms that are also special cases---Thread-Greedy CD and Coloring-Based CD---and give performance measurements for an OpenMP implementation of these.

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。

↑