Causal Inference For Socio-Economic & Engineering Systems

on YouTube ↗11m

" summary

Work bridges econometric causal methods and ML by treating counterfactuals as tensor/matrix completion problems, using synthetic interventions to impute missing treatment trajectories and proving entrywise finite-sample consistency and asymptotic normality under a low-rank assumption. Applied to a failed Parex Alzheimer's trial the procedure identifies 20% responsive patients and informs a new trial; an Uber study predicts larger US cities see increased wait times. Algorithmic tool is synthetic nearest neighbors for robust, local completion.

AI-written summary · about

" chapters (8)

1. 0:00Key framing question
2. 0:23What will happen to Y if we do A?
3. 0:53Research agenda: bridging causal inference/econometrics & machine learning
4. 2:38Synthetic interventions (SI): counterfactual reasoning via tensor completion
5. 3:51Same questions, different language
6. 4:32Synthetic interventions: widely applied across many social & engineering systems
7. 6:16Theoretical results
8. 7:50Building causal simulators for RL

" related talks

• 30mCausal Inference For Socio-Economic And Engineering SystemsAnish Agarwal · Learning from Interventions 2022
• 30mKernel Methods For Causal InferenceRahul Singh · Algorithmic Aspects of Causal Inference 2022
• 47mCausal Matrix Completion: Applications to Offline Causal Reinforcement LearningAnish Agarwal · Quantifying Uncertainty: Stochastic, Adversarial, and Beyond
• 29mCausal Matrix CompletionAnish Agarwal · Algorithmic Aspects of Causal Inference 2022
• 01h 01mThe Mathematics of Causal Inference, with Reflections on Machine Learning and the Logic of ScienceJudea Pearl · Symposium on Visions of the Theory of Computing 2013
• 06mCausal Machine Learning – What’s in it for Data Scientists?Adi Watzman · PyData Tel Aviv 2022
• 31mA Gentle Introduction To Causal InferenceEyal Kazin · PyData Global 2022
• 29m🧠🧹 Causality - Mental Hygiene for Data ScienceEyal Kazin · PyData Global 2024