Dependence of variance on covariate design in nonparametric link regression
Abstract
This paper discusses a design-dependent nature of variance in nonparametric link regression aiming at predicting a mean outcome at a link, i.e., a pair of nodes, based on currently observed data comprising covariates at nodes and outcomes at links.
- Publication:
-
arXiv e-prints
- Pub Date:
- December 2020
- DOI:
- 10.48550/arXiv.2012.13106
- arXiv:
- arXiv:2012.13106
- Bibcode:
- 2020arXiv201213106O
- Keywords:
-
- Mathematics - Statistics Theory
- E-Print:
- 24 pages, 11 figures, accepted to Statistics and Probability Letters