Abstract:This paper concerns model averaging for generalized partially linear single-index models(GPLSIMs).GPLSIMs have attracted increasing attention due to their great flexibility and easy interpretability.However,it is tough to determine the predictors and the degree of smoothing in the single-index link function in practice.To deal with this model selection uncertainty problem,we propose an optimal model averaging approach for GPLSIMs,in which we select the data-driven weights by maximizing a cross-validation criterion.Under model misspecification,we prove that the resulting model averaging estimator is asymptotically optimal in the sense of achieving the lowest possible Kullback-Leibler loss in the framework of diverging model spaces.On the other hand,when pseudo-correct models are available in the candidate model set,we demonstrate that the cross-validation scheme asymptotically assigns all weights to the pseudo-correct models.Moreover,based on our model averaging approach,we develop a reliable variable importance measure for GPLSIMs,which can be shown to asymptotically identify all the variables in the true model.Simulation experiments and two empirical applications illustrate the superiority of the proposed method over several existing methods.

Optimal Model Averaging for Semiparametric Partially Linear Models with Measurement Errors

Optimal Model Averaging for Semiparametric Partially Linear Models with Censored Data

Model Selection and Model Averaging for Semiparametric Partially Linear Models with Missing Data

Optimal Model Averaging Estimator for Semi-Functional Partially Linear Models

Optimal model averaging for partially linear models with missing response variables and error‐prone covariates

Model Averaging Estimation for Nonparametric Varying-Coefficient Models with Multiplicative Heteroscedasticity

Model averaging estimation for generalized partially linear varying-coefficient models

Model averaging for right censored data with measurement error

Model Averaging for Semiparametric Varying Coefficient Quantile Regression Models

Penalized Time-Varying Model Averaging

Optimal Model Averaging Estimation for Generalized Linear Models and Generalized Linear Mixed-Effects Models

A Mallows-type Model Averaging Estimator for De-Noise Linear Models

Model Averaging Estimation for Partially Linear Functional Score Models

Focused Information Criterion and Model Averaging for Varying-Coefficient Partially Linear Models with Longitudinal Data

Partial Linear Model Averaging Prediction for Longitudinal Data

Model Averaging for Varying Coefficient Partially Linear Models With Missing Data

Optimal Model Averaging by Data Perturbation for Spatial Noisy Data

Model Averaging Procedure for Varying-Coefficient Partially Linear Models with Missing Responses

Optimal Model Averaging for Generalized Partially Linear Single-Index Models

Model Averaging for Estimating Treatment Effects With Binary Responses

Model Averaging with Averaging Covariance Matrix