arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2008.06296stat.MLcs.LGstat.AP
收藏

高维最小二乘估计中可证明的较多数据损伤

Provable More Data Hurt in High Dimensional Least Squares Estimator

Southern University of Science and Technology(南方科技大学);Fudan University(复旦大学)

Zeng Li, Chuanlong Xie, Qinwen Wang

摘要:

本文研究了高维最小二乘估计的有限样本预测风险。当样本量和特征数都趋于无穷大时,我们推导出了预测风险的中心极限定理。给出了预测风险的有限样本分布和置信区间。我们的理论结果证明了预测风险的样本非单调性,并证实了“更多的数据伤害”现象。

英文摘要:

This paper investigates the finite-sample prediction risk of the high-dimensional least squares estimator. We derive the central limit theorem for the prediction risk when both the sample size and the number of features tend to infinity. Furthermore, the finite-sample distribution and the confidence interval of the prediction risk are provided. Our theoretical results demonstrate the sample-wise nonmonotonicity of the prediction risk and confirm "more data hurt" phenomenon.

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。

↑