出版社：Suntory Toyota International Centre for Economics and Related Disciplines
摘要：The nonparametric censored regression model, with a fixed, known censoringpoint (normalized to zero), is y = max[0,m(x) + e], where both the regressionfunction m(x) and the distribution of the error e are unknown. This paperprovides estimators of m(x) and its derivatives. The convergence rate is thesame as for an uncensored nonparametric regression and its derivatives. Wealso provide root n estimates of weighted average derivatives of m(x), whichequal the coefficients in linear or partly linearr specifications for m(x). Anextension permits estimation in the presence of a general form ofheteroscedasticity. We also extend the estimator to the nonparametrictruncated regression model, in which only uncensored data points areobserved. The estimators are based on the relationship.E(yk\x)/.m(x) = kE[yk-1/(y > 0)x ], which we show holds for positive integers k
