|
Activity Number:
|
386
- Nonparametric Modeling II
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 12, 2021 : 12:00 PM to 1:50 PM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
|
Abstract #317895
|
|
|
Title:
|
Endogeneity in Modal Regression
|
|
Author(s):
|
Tao Wang*
|
|
Companies:
|
University of California, Riverside
|
|
Keywords:
|
Conditional mode independence;
Control function;
Endogeneity;
Modal regression;
Semiparametric partially linear;
Variable selection
|
|
Abstract:
|
In this paper, we propose a control function approach to account for endogeneity in a parametric linear triangular simultaneous equations model for modal regression when the conditional mode of unobservable error term on explanatory variables is nonzero. We adjust endogeneity with the residuals from the conditional mode decomposition of the endogenous variable as controls in the structural equation, and develop a computationally attractive two-step estimation procedure with the conditional mode independence restriction. The proposed estimators could be easily solved by virtue of a modified modal expectation-maximization (MEM) algorithm. Consistency and asymptotic properties of the estimators for both parametric and nonparametric parts are rigorously established under generic regularity conditions. Monte Carlo simulations are conducted to examine the finite sample performance of the proposed estimation procedures. Two applications to the real datasets are presented to further illustrate the proposed estimators in practice. We in the end develop an adaptive lasso method to select instrumental variables and demonstrate the oracle property of the proposed penalized modal regression.
|
Authors who are presenting talks have a * after their name.
