蜜桃传媒app

Leila Golparvar, PhD

Postdoctoral Researcher, Department of Mathematics and Statistics, McGill University

Causal Structure Learning and Propensity Score Adjustment

Joint work with: D.A. Stephens and R. Platt

ALL ARE WELCOME

Abstract:

Mathematical representation of causal dependencies among a set of variables via graphs has been in the statistical literature for more than a century. A graph G = (V , E) consists of a vertex set V = {X1,...,Xp} representing observed variables, and an edge set E representing structural links between the variables. Predict颅ing the e铿�ect of manipulations from non-experimental data therefore often involves two steps: 铿乺st, discovery of causal structures, represented by directed acyclic graphs (DAGs), and second identi铿乧ation and estimation of causal parameters.

In this talk, we will adopt a frequentist constraint-based approach to discover the causal graph, and use the PC algorithm for causal discovery. The PC algorithm is based on sequential testing of partial correlations between variables and a search strategy that identi铿乪s the presence and absence of structural links. To explore the use of the PC algorithm in selection of confounders, we conduct a Monte Carlo simulation study. It is known that adjusting for all confounders will eliminate bias, but additionally adjusting for predictors of outcome that are unrelated to treatment will lead to estimates with lower variance. Results show that PC algorithm works very well when the sample size is moderate to large.

KEYWORDS: Causality; Confounding; Counterfactual; Directed acyclic graph; PC-algorithm.

听

听

听