Logarithmic calibration for nonparametric multiplicative distortion measurement errors models

A logarithmic calibration estimation procedure is proposed for nonparametric regression models under the multiplicative distortion measurement errors setting. The unobservable response variable and covariates are both distorted in a fashion by an observed confounding variable. By using unobserved variables, we consider to study estimates of mean function its first derivative, variance function, Sharpe ratio correlation curve. We obtain asymptotic normality results estimators. Monte Carlo simulation experiments conducted examine performance estimators applied analyse real dataset illustration.