Research Based on High-Dimensional Fused Lasso Partially Linear Model

In this paper, a partially linear model based on the fused lasso method is proposed to solve problem of high correlation between adjacent variables, and then idea two-stage estimation used study solution model. Firstly, non-parametric part estimated using kernel function transforming semiparametric into parametric Secondly, regularization term introduced construct least squares parameter penalty. Then, due non-smooth terms model, subproblems may not have closed-form solutions, so linearized alternating direction multiplier (LADMM) convergence algorithm asymptotic properties are analyzed. Finally, applicability was demonstrated through two types simulation data practical problems in predicting worker wages.