B0487
Title: KOO: A scalable model selection rules in high-dimensional regression
Authors: Zhidong Bai - Northeast Normal University (China)
Kwok Pui Choi - National University of Singapore (Singapore)
Yasunori Fujikoshi - Hiroshima university (Japan)
Jiang Hu - Northeast Normal University (China) [presenting]
Abstract: Variable selection is essential for improving inference and interpretation in multivariate linear regression. We consider the strong consistency of the high-dimensional knock-one-out (KOO) methods. The proposed method removes the penalties while simultaneously reducing the conditions for the dimensions and sizes of the regressors. Simulation studies and real data analysis support our conclusions.
