M. I. Borrajo García, W. González Manteiga, M. D. Martínez Miranda
The first-order intensity function is one of the characteristic functions of a point process and has generated a great interest since the eighties. Allowing for covariate dependence agrees to gather extra information and thus to improve the estimation of the intensity function.
This work provides with a new kernel intensity estimator with covariates, for which a complete theoretical framework to guarantee its consistency is developed. A new bootstrap resampling procedure, as well as two new specific data-driven bandwidth selection methods are also defined. Moreover a L2-distance based test statistic is proposed to check whether the assumed model with covariates for the intensity function is or not appropriate. The asymptotic normality of the test statistic is derived and the calibration of the test is improved through a bootstrap method. The good behaviour of all the proposed techniques is shown in different simulation studies and in applications to real data sets...
Palabras clave / Keywords: ...
Programado
Sesión 1 Premio Ramiro Melendreras
29 de mayo de 2018 09:10
Sala Cristal
