arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 1208.0129stat.MLcs.LG
收藏

计算自适应模型选择的预言不等式

Oracle inequalities for computationally adaptive model selection

University of California, Berkeley(加州大学伯克利分校);Queensland University of Technology(昆士兰科技大学)

Alekh Agarwal, Peter L. Bartlett, John C. Duchi

摘要:

我们分析了一般的模型选择过程中使用惩罚的经验损失最小化计算约束。虽然经典的模型选择方法不考虑计算方面的执行模型选择,我们认为,任何实际的模型选择程序不仅必须权衡估计和近似误差,但也需要计算经验最小值为不同的功能类的计算工作。我们提供了一个框架来分析这样的问题,我们给模型选择下的计算预算的算法。这些算法满足预言不等式,表明所选模型的风险并不比我们将所有的计算预算都投入到最优函数类中差得多。

英文摘要:

We analyze general model selection procedures using penalized empirical loss minimization under computational constraints. While classical model selection approaches do not consider computational aspects of performing model selection, we argue that any practical model selection procedure must not only trade off estimation and approximation error, but also the computational effort required to compute empirical minimizers for different function classes. We provide a framework for analyzing such problems, and we give algorithms for model selection under a computational budget. These algorithms satisfy oracle inequalities that show that the risk of the selected model is not much worse than if we had devoted all of our omputational budget to the optimal function class.

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。

↑