Nonparametric relative regression estimation for associated and randomly left truncated model

par Farida Hamrani (INSA Toulouse)

mardi 10 juin 2025, 11:15 → 12:15 Europe/Paris

Salle K. Johnson (1R3, 1er étage)

We consider the problem of estimating non-parametrically the regression function when the target random variable is subject to random left truncation. The estimator we deal with here is based on the minimization of the mean squared relative error. As a main result, we state the uniform almost sure convergence with rate of this estimator under association dependency. The performance of the studied estimator is evaluated via a simulation study and is compared with that proposed by Ould Saïd and Lemdani (2006).

Keywords and phrases: Association, Kernel regression estimator,
Random left-truncation model, Relative error, Strong consistency