About: Synthetic Interventions

Facets (new session)
Description
Metadata
Settings
- owl:sameAs
- Inference Rule:

About: Synthetic Interventions Goto Sponge NotDistinct Permalink

An Entity of Type : schema:ScholarlyArticle, within Data Space : covidontheweb.inria.fr associated with source document(s)

Attributes	Values
type	Academic Article research paper schema:ScholarlyArticle
isDefinedBy	Covid-on-the-Web dataset
has title	Synthetic Interventions
Creator	Shen, Agarwal, Alomar Agarwal, Anish Agarwal, C Alomar, Abdullah Alomar, A Cosson, R Cosson, Romain Cosson, Shah Shah, Devavrat Shah, D Shen, Dennis Shen, &
Source	ArXiv
abstract	We develop a method to help quantify the impact different levels of mobility restrictions could have had on COVID-19 related deaths across nations. Synthetic control (SC) has emerged as a standard tool in such scenarios to produce counterfactual estimates if a particular intervention had not occurred, using just observational data. However, it remains an important open problem of how to extend SC to obtain counterfactual estimates if a particular intervention had occurred - this is exactly the question of the impact of mobility restrictions stated above. As our main contribution, we introduce synthetic interventions (SI), which helps resolve this open problem by allowing one to produce counterfactual estimates if there are multiple interventions of interest. We prove SI produces consistent counterfactual estimates under a tensor factor model. Our finite sample analysis shows the test error decays as $1/T_0$, where $T_0$ is the amount of observed pre-intervention data. As a special case, this improves upon the $1//sqrt{T_0}$ bound on test error for SC in prior works. Our test error bound holds under a certain%22subspace inclusion%22condition; we furnish a data-driven hypothesis test with provable guarantees to check for this condition. This also provides a quantitative hypothesis test for when to use SC, currently absent in the literature. Technically, we establish the parameter estimation and test error for Principal Component Regression (a key subroutine in SI and several SC variants) under the setting of error-in-variable regression decays as $1/T_0$, where $T_0$ is the number of samples observed; this improves the best prior test error bound of $1//sqrt{T_0}$. In addition to the COVID-19 case study, we show how SI can be used to run data-efficient, personalized randomized control trials using real data from a large e-commerce website and a large developmental economics study.
has issue date	2020-06-13(xsd:dateTime)
has license	arxiv
sha1sum (hex)	0d55a450f1fc696a73ac5d3477617c1c9b9a259f
resource representing a document's title	Synthetic Interventions
resource representing a document's body	covid:0d55a450f1fc696a73ac5d3477617c1c9b9a259f#body_text
is schema:about of	named entity 'e-commerce' named entity 'high-dimensional' named entity 'parameter estimation' named entity 'out-of-sample' named entity 'counterfactual' named entity 'error bound' named entity 'case study' named entity 'factor model' named entity 'triangle inequality' named entity 'gaussian' named entity 'PCR' named entity 'probability' named entity 'covariate' named entity 'probability' named entity 'random vectors' named entity 'Var' named entity 'conditional independence' named entity 'null space' named entity '0.98' named entity '2.2' named entity 'vector' named entity 'supervised learning' named entity 'infection' named entity 'COVID-19 pandemic' named entity 'e-commerce company' named entity 'gaussian' named entity 'Lemma 28' named entity 'transductive learning' named entity 'Gaussian coordinates' named entity 'error bound' named entity 'gaussian' named entity 'linearity of expectations' named entity 'multivariate gaussian' named entity 'factor model' named entity 'e-commerce' named entity 'Parameter Estimation' named entity '2.2' named entity 'quadratic form' named entity 'linear algebraic' named entity 'Bernoulli random variables' named entity '2.2' named entity 'probability' named entity 'hypothesis test' named entity 'supremum' named entity 'Principal Component Regression' named entity 'probability' named entity 'data generating process' named entity 'Lemma 28' named entity 'covariate' named entity 'PCR' named entity 'potential outcomes' named entity 'i.i.d' named entity 'Hölder's inequality' named entity 'India' named entity 'target country' named entity 'death trajectory' named entity 'Hypothesis Test' named entity 'PCR' named entity 'error bound' named entity 'order-three tensor' named entity 'latent factors' named entity 'randomized control trials' named entity 'rank tensor' named entity 'factor model' named entity 'linear model' named entity 'significance level' named entity 'convex set' named entity 'observational data' named entity 'real-world' named entity 'probability' named entity 'COVID-19' named entity 'counterfactual' named entity 'PCA' named entity 'factor model' named entity 'test statistic'

Faceted Search & Find service v1.13.91 as of Mar 24 2020

Alternative Linked Data Documents: Sponger | ODE Content Formats:

RDF

ODATA

Microdata

About

OpenLink Virtuoso version 07.20.3229 as of Jul 10 2020, on Linux (x86_64-pc-linux-gnu), Single-Server Edition (94 GB total memory)
Data on this page belongs to its respective rights holders.
Virtuoso Faceted Browser Copyright © 2009-2024 OpenLink Software