Variable Selection for Qualitative Interactions

By Lacey Gunter, Ji Zhu, Susan A. Murphy. 2011

86 citations recorded by the source corpus.

In this article we discuss variable selection for decision making with focus on decisions regarding when to provide treatment and which treatment to provide. Current variable selection techniques were developed for use in a supervised learning setting where the goal is prediction of the response. These techniques often downplay the importance of interaction variables that have small predictive ability but that are critical when the ultimate goal is decision making rather than prediction. We propose two new techniques designed specifically to find variables that aid in decision making. Simulation results are given along with an application of the methods on data from a randomized controlled trial for the treatment of depression.

Subfields: Experimental and Cognitive Psychology, Statistics and Probability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Experimental and Cognitive Psychology

In Statistics and Probability

Citations in the corpus

Residual Weighted Learning for Estimating Individualized Treatment Rules (2015, background): “It has been noticed in practical settings that the main effects, rather than treatment-covariate interactions, tend to explain most of the variability in the outcome, and thus they are important for good prediction (Gunter et al. 2011).”
Boosting predictions of treatment success (2010, methodology): “These proposed statistical models would go beyond simple prediction to be prescriptive (13), because such models”
On Optimal Treatment Regimes Selection for Mean Survival Time (2014, unknown): “For example, [13] introduced the concepts for predictive variables (predictors with important baseline effects) and prescriptive variables (predictors interacting with treatments) and developed a ranking method for selecting qualitative interactions.”
Precision Medicine. (2019, background): “…has been studied as a subtopic within precision medicine with early approaches seeking to identify ‘qualitative interactions’ (Gail and Simon 1985; Gunter et al 2011) and more recent approaches focused on identification of variables that are informative for identification of an optimal…”
Tools for the Precision Medicine Era: How to Develop Highly Personalized Treatment Recommendations From Cohort and Registry Data Using Q-Learning (2017, methodology): “(14)) could be used to eliminate unnecessary variables, although this technique has not been adopted due to computational complexity.”
Budgeted Learning for Developing Personalized Treatment (2014, methodology): “[23] suggested 2 potential variables–Obsessive Compulsiveness (OC) and Alcohol–for making treatment decisions.”
Estimating Optimal Treatment Regimes from a Classification Perspective (2012, background): “, an interaction in which the treatment effect changes direction depending on the covariates (Gunter et al., 2011).”