Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 143 - New Machine Learning Tools for Mobile Health Data and Individual Intervention
Type: Invited
Date/Time: Tuesday, August 10, 2021 : 10:00 AM to 11:50 AM
Sponsor: Lifetime Data Science Section
Abstract #315514
Title: Inference of Causal Relations with Interventions
Author(s): Chunlin Li and Xiaotong T Shen* and Wei Pan
Companies: University of Minnesota and University of Minnesota and University of Minnesota
Keywords: High-dimensional inference; identifiability; testability; Likelihood ratio tests

The inference of causal relations between primary variables is challenging in the presence of unknown interventions. In this article, we infer multiple causal relations while identifying relevant interventions. In particular, we derive conditions for multiple unknown interventions to yield an identifiable model. For inference, we need to identify the ancestral relations and the interventions for each hypothesis-specific primary variable. Towards this end, we propose a causal discoveryalgorithm. On this ground, we propose a likelihood ratio test based on data perturbation, in which the identification effect is accounted for by perturbing original data to assess the uncertainty associated with identifying ancestors and interventions. For testing the presence and strengths of causal relations in a pathway, we show that the proposed tests achieve desired statistical properties in terms of controlling Types I and II error in a higher-dimensional situation. Numerical examples will be given to demonstrate the utility and effectiveness of the proposed procedure.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2021 program