Causal Modelling of Heavy-Tailed Variables and Confounders with Application to River Flow

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Pasche, Olivier C., Chavez-Demoulin, Valérie, Davison, Anthony C.
Format: Preprint
Veröffentlicht: 2021
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929589289222144
author Pasche, Olivier C.
Chavez-Demoulin, Valérie
Davison, Anthony C.
author_facet Pasche, Olivier C.
Chavez-Demoulin, Valérie
Davison, Anthony C.
contents Confounding variables are a recurrent challenge for causal discovery and inference. In many situations, complex causal mechanisms only manifest themselves in extreme events, or take simpler forms in the extremes. Stimulated by data on extreme river flows and precipitation, we introduce a new causal discovery methodology for heavy-tailed variables that allows the effect of a known potential confounder to be almost entirely removed when the variables have comparable tails, and also decreases it sufficiently to enable correct causal inference when the confounder has a heavier tail. We also introduce a new parametric estimator for the existing causal tail coefficient and a permutation test. Simulations show that the methods work well and the ideas are applied to the motivating dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2110_06686
institution arXiv
publishDate 2021
record_format arxiv
spellingShingle Causal Modelling of Heavy-Tailed Variables and Confounders with Application to River Flow
Pasche, Olivier C.
Chavez-Demoulin, Valérie
Davison, Anthony C.
Methodology
Applications
62G32
Confounding variables are a recurrent challenge for causal discovery and inference. In many situations, complex causal mechanisms only manifest themselves in extreme events, or take simpler forms in the extremes. Stimulated by data on extreme river flows and precipitation, we introduce a new causal discovery methodology for heavy-tailed variables that allows the effect of a known potential confounder to be almost entirely removed when the variables have comparable tails, and also decreases it sufficiently to enable correct causal inference when the confounder has a heavier tail. We also introduce a new parametric estimator for the existing causal tail coefficient and a permutation test. Simulations show that the methods work well and the ideas are applied to the motivating dataset.
title Causal Modelling of Heavy-Tailed Variables and Confounders with Application to River Flow
topic Methodology
Applications
62G32
url https://arxiv.org/abs/2110.06686