A new partitioning criterion based on kernel density estimates of covariates achieves better balance and more accurate difference-in-mean estimators than complete randomization or rerandomization in controlled experiments.
Two-sam ple test statistics for measuring discrepancies between two multivariate probability de nsity functions using kernel-based density estimates
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2020 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Covariate Balancing Based on Kernel Density Estimates for Controlled Experiments
A new partitioning criterion based on kernel density estimates of covariates achieves better balance and more accurate difference-in-mean estimators than complete randomization or rerandomization in controlled experiments.