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 1502.03175stat.MLcs.LGstat.ME
收藏

统计与机器学习中的近似算法

Proximal Algorithms in Statistics and Machine Learning

University of Chicago(芝加哥大学);University of Texas at Austin(得克萨斯大学奥斯汀分校)

Nicholas G. Polson, James G. Scott, Brandon T. Willard

摘要:

在本文中,我们开发了统计学习的近似方法。邻近点算法在统计学和机器学习中用于获得复合函数的优化解是有用的。我们的方法利用封闭形式的解决方案的近端运营商和包络表示的基础上莫罗,向前向后,道格拉斯-Rachford和半二次包络。包络表示导致新的近端算法的复合目标函数,其中包括非光滑和非凸目标的统计优化。我们用正则化的Logistic和Poisson回归以及具有融合套索范数的非凸桥惩罚来说明我们的方法。我们提供了一个讨论的收敛性的非下降算法的加速和非凸函数。最后,我们为未来的研究提供了方向。

英文摘要:

In this paper we develop proximal methods for statistical learning. Proximal point algorithms are useful in statistics and machine learning for obtaining optimization solutions for composite functions. Our approach exploits closed-form solutions of proximal operators and envelope representations based on the Moreau, Forward-Backward, Douglas-Rachford and Half-Quadratic envelopes. Envelope representations lead to novel proximal algorithms for statistical optimisation of composite objective functions which include both non-smooth and non-convex objectives. We illustrate our methodology with regularized Logistic and Poisson regression and non-convex bridge penalties with a fused lasso norm. We provide a discussion of convergence of non-descent algorithms with acceleration and for non-convex functions. Finally, we provide directions for future research.

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。

↑