- Thesis:
- Bachelor in Mathematics
- Author:
- Lena Krienke
- Title:
- Calibrating data-driven estimators in presence of censoring
- Supervisors:
- Sergio Brenner Miguel and Jan JOHANNES
- Abstract:
- In this bachelor thesis we focus on the nonparametric estimation of a density f of a positive real-valued random variable X in the multiplicative censoring model. Instead of observing the random variable X we take a look on the random variable Y which is the product of X and U, hence Y = XU, with U as a β(1,k)-distributed random variable and X and U are independent. We project the density f on a subspace which is the linear span of the first m functions of the Laguerre basis. The projected density fm is estimated by fˆ . Then we get assertions about the mean integrated squared error of the density f and the projected density fˆ regarding to the upper and lower bound. Since fˆ primarily depends on m we have to estimate this parameter by a data-driven estimator. Subsequently we accentuate the theoretical statements by a Monte-Carlo simulation and have a view on various factors like the effect of an increasing k and different sample sizes.
- References:
- Belomestny, Comte and Genon-Catalot. Nonparametric Laguerre estimation in the multiplicative censoring model, Electronic Journal of Statistics, 10(2):3114–315, 2016.
- Brunel, Comte and Genon-Catalot: Nonparametric density and survival function estimation in the multiplicative censoring model, Test 25(3):570–590, 2016.
| Statistics Group | Institute for Mathematics | Faculty of Mathematics and Computer Science | University Heidelberg |