Скачать или смотреть Causal and non-causal paths

Episode C4VL1: Causal and non-causal paths.

In this episode of 𝗖𝗮𝘂𝘀𝗮𝗹𝗶𝘁𝘆 𝘄𝗶𝘁𝗵 𝗕𝗼𝗿𝗶𝘀 you will learn that all associations are produced by paths, but not all paths produce associations. Causal inference arises from reasoning about open and closed paths of causal diagrams.

0:00 Intro
0:35 Identification problem
3:49 A general plan to tackle it
6:20 Associations produced by paths
8:16 Blocked and open paths
11:13 Take-aways

Links to articles:
• Miettinen, Stratification: American Journal of Epidemiology, Volume 104, Issue 6, December 1976: 609–620, https://tinyurl.com/miettinen
• d-separation in DAGitty https://tinyurl.com/DAGgitty
• Pearl on d-separation https://tinyurl.com/withNOtears

FOLLOW ME ON TWITTER @soboleffspaces
YOU MAY ENJOY MY BLOG www.sobolevspace.com
NEWSLETTER https://world.hey.com/aiia/class-4-vi...

#causality #CausalInference #CausalPaths
#CausalReasoning #CausalAnalysis #CausalDiscovery
#Identification #GraphicalCriteria #dseparation
#CausalPaths #NoncausalPaths #DependencyGraph
#TreatmentEffect #conditioning #associations
#openpaths #CausalDiagram #NoncausalPaths
#DirectedAcyclicGraph
#CausalAttribution #MachineLearning #CausalDataAnalysis
#sobolevspaces