Estimating a monotone probability mass function with known flat regions
Statistics Theory
2016-12-13 v4 Statistics Theory
Abstract
We propose a new estimator of a discrete monotone probability mass function with known flat regions. We analyse its asymptotic properties and compare its performance to the Grenander estimator and to the monotone rearrangement estimator.
Cite
@article{arxiv.1612.00196,
title = {Estimating a monotone probability mass function with known flat regions},
author = {Dragi Anevski and Vladimir M. Pastukhov},
journal= {arXiv preprint arXiv:1612.00196},
year = {2016}
}
Comments
16 pages, 1 figure