MPL.control

Estimated propentisy score for each patient. If not specified, a 
 logistic regression model is fitted to estimate the propensity score.

pi.est

Estimated baseline function for each patient. If not specified, a linear
 regression model is fitted to estimate the baseline function.

h.est

Number of bootstrap samples used for inference of the maximin effects
 and the subgroup parameter. Default is 600.

boot.sample

Parameters that control fitting of maximin projection learning.

Maximin-projection learning (MPL, Shi, et al., 2018) is implemented for
recommending a meaningful and reliable individualized treatment regime for future
groups of patients based on the observed data from different populations with
heterogeneity in individualized decision making. Q-learning and A-learning are
implemented for estimating the groupwise contrast function that shares the same
marginal treatment effects. The packages contains classical Q-learning and A-learning
algorithms for a single stage study as a byproduct. More functions will be added
at later versions.

MPL.control: Control parameters for maximin projection learning

Description

Usage

Arguments

Value

See Also