Study bulletin: residuals density estimation in censored linear linear regression model

摘要

Suppose that in the linear regression model Y_i=X_iβ+e_i. Y_i are not completely observable and observations are(X_i,Z_i,δ_i), i=1,...,n, whereβand X_i and p×1 vectors, Z_i=min(T_i,W_i),δ_i=I(Y_i ≤W_i), the residual e_i's are i.i.d.r.v.'s with unknown common density function f(distribution function F)and the(W_i,X_i)'s are i.i.d. random vectors that are independent of e_i's. Fitting the model with censored data has received considerable attention in the statis- Tical literature.