Youtor
Causality part 1 Bernhard Schölkopf MLSS 2020 Tübingen
>> YOUR LINK HERE: ___ http://youtube.com/watch?v=btmJtThWmhA
Table of Contents (powered by https://videoken.com) • 0:00:00 Causality, part 1 - Bernhard Scholkopf - MLSS 2020, Tubingen • 0:00:41 Introduction to the Speaker • 0:01:52 Causality • 0:02:27 Roadmap • 0:05:05 Notation • 0:06:46 Independence • 0:07:35 Independence of random variables • 0:09:19 Conditional Independence of random variables • 0:10:31 What is cause and what is effect? • 0:14:09 (Physical) independence of mechanisms • 0:16:15 Reichenbach's Common Cause Principle • 0:17:34 Definition of a Structural Causal Model (Pearl et al.) • 0:19:46 Reichenbach's Principle and causal sufficiency • 0:21:26 Entailed distribution • 0:24:16 Markov conditions • 0:27:39 Graphical Causal Inference (Spirtes, Glymour, Scheines, Pearl, ...) • 0:29:58 Interventions and shifts • 0:32:21 Independent mechanisms and disentangled factorizations • 0:39:12 Counterfactuals • 0:41:29 Does it make sense to talk about statistics without mentioning time? • 0:43:16 Causality in differential equations • 0:45:03 A Modeling Taxonomy • 0:46:28 From Ordinary Differential Equations to Structural Causal models for the deterministic case • 0:47:35 Recap • 0:47:54 Pearl's do calculus • 0:49:01 Difference between seeing and doing • 0:50:00 Computing p(X1, . . ., Xn\\doxi) • 0:50:55 Computing p(Xk\\do xi) • 0:52:06 Examples for p(.\\dox) not equal to p(.\\x) • 0:52:36 Controlling for confounding / adjustment formula • 0:57:47 Simpson's paradox in Covid-19 case fatality rates • 1:00:22 Coarse-grained causal graph • 1:00:28 Mediation analysis • 1:01:35 Recap: Structural Causal Model • 1:01:44 Twilight of the Idols • 1:02:19 Restricting the Structural Causal Model • 1:04:24 Causal Inference with Additive Noise, 2-Variable Case • 1:05:19 Intuition • 1:06:01 Alternative View • 1:06:03 Causal Inference Method • 1:06:17 Experiments • 1:07:02 Independence-based Regression • 1:07:19 Causal Inference Method • 1:07:51 Side note on multivariate ANMs • 1:07:52 Independence-based Regression • 1:08:27 Independence of input and mechanism • 1:09:29 Inferring deterministic causal relations • 1:10:57 Causal independence implies anticausal dependence • 1:11:47 Benchmark dataset with 106 cause-effect pairs • 1:11:57 Cause-Effect Pairs - Examples • 1:12:36 Causal Learning and Anticausal Learning • 1:14:12 Covariate Shift and Semi-Supervised Learning • 1:15:58 Experimental Meta-Analysis confirms prediction • 1:16:28 Higher-order Semi-Supervised Learning • 1:16:58 Algorithmic structural causal model • 1:18:03 Gedankenexperiment • 1:18:21 Thermodynamic Arrow of Time • 1:18:39 Milky Way Galaxy • 1:21:00 Half-Sibling Regression • 1:21:42 Planet-Hunting Kepler Spacecraft Suffers Major Failure, NASA Says • 1:22:52 Habitable Zone Gallery • 1:24:01 Outlook: Causal discovery (Spirtes, Glymour, Scheines, Pear) • 1:24:09 Outlook: Causal representation learning • 1:24:54 Causal mechanisms in machine learning • 1:25:01 End of Part 1 • 1:25:14 Q A
#############################
