IntDC Framework Identifies Intervention-Level Causality From Observational Data Alone

Written by

in

The research team introduces the interventional dynamic causal framework, IntDC, which uses interventional embedding entropy (IEE) to quantify causal information between variables from observational time-series data, without requiring actual interventions or prior knowledge of the system’s dynamical equations. The method reduces spurious associations caused by confounding in tests on classical dynamical systems and chaotic neural networks, and has been applied to neural connectivity in C. elegans, the regional spread of COVID-19, and circadian gene regulatory networks. The paper was published in The Innovation, with DOI 10.1016/j.xinn.2026.101358.

Source:Site Inovação Tecnológica · Read Original

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *