New efficient estimation and variable selection in models with single-index structure

作者：Wang Kangning^*; Lin Lu

来源：Statistics & Probability Letters, 2014, 89: 58-64.

DOI：10.1016/j.spl.2014.02.019

• Linearity condition
• Composite quantile regression
• Adaptive LASSO
• Oracle property
• Single-index structure

We propose a new efficient root-n consistent estimate for the direction of the index parameter vector in a class of models with single-index structure. To select the important predictors, we suggest using the adaptive LASSO and establish the oracle property. Simulation results also confirm the theoretical findings.

• 出版日期2014-6
• 单位山东大学; 重庆文理学院