Counterfactual Inference For Sequential Experiment Design

on YouTube ↗30m

" summary

A policy-agnostic nearest-neighbor estimator treats each treatment as an N-by-T matrix and reduces counterfactuals to sequential matrix completion, proving consistency under a bilinear latent-factor model and a central limit theorem (normal fluctuations at rate sqrt(N_neighbors)) for Lipschitz nonlinearity at the final time. Requirements are lower-bounded sampling probabilities, diverse time factors, non-adversarial noise and slow adaptive changes (a small gamma term). Tuning uses cross-validation and a preprint is on arXiv.

AI-written summary · about

" chapters (8)

1. 0:00Introduction
2. 1:23Setup
3. 4:13Outline
4. 7:09Reduction
5. 8:17Approximation
6. 10:42Latent Factor Model
7. 20:30Central Limit Theorem
8. 23:55Expected squared error bound

" related talks

• 47mCausal Matrix Completion: Applications to Offline Causal Reinforcement LearningAnish Agarwal · Quantifying Uncertainty: Stochastic, Adversarial, and Beyond
• 33mSimple Yet Efficient Estimators For Network Causal Inference...Christina Yu · Algorithmic Aspects of Causal Inference 2022
• 29mCausal Matrix CompletionAnish Agarwal · Algorithmic Aspects of Causal Inference 2022
• 11mCausal Inference For Socio-Economic & Engineering SystemsAnish Agarwal · Meet the Fellows Welcome Event
• 49mCausal Inference in Complex Systems: Network Interference, Strategic Agents, and BeyondPanos Toulis · Quantifying Uncertainty: Stochastic, Adversarial, and Beyond
• 52mCausal Matrix EstimationDevavrat Shah · Algorithmic Advances for Statistical Inference with Combinatorial Structure
• 39mModern Bayesian Experimental DesignDesi R. Ivavona · LAFI 2025
• 30mNon-parametric Causal Inference in Dynamic Thresholding DesignsStefan Wager · Bridging Prediction and Intervention Problems in Social Systems (Simons Institute, 2026)