Using Deep Autoregressive Models as Causal Inference Engines

Existing causal inference (CI) models are limited to primarily handling low-dimensional confounders and singleton actions. We propose an autoregressive (AR) CI framework capable of handling complex confounders and sequential actions common in modern applications. We accomplish this by {\em sequencif...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Im, Daniel Jiwoong, Zhang, Kevin, Verma, Nakul, Cho, Kyunghyun
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Im, Daniel Jiwoong
Zhang, Kevin
Verma, Nakul
Cho, Kyunghyun
description Existing causal inference (CI) models are limited to primarily handling low-dimensional confounders and singleton actions. We propose an autoregressive (AR) CI framework capable of handling complex confounders and sequential actions common in modern applications. We accomplish this by {\em sequencification}, transforming data from an underlying causal diagram into a sequence of tokens. This approach not only enables training with data generated from any DAG but also extends existing CI capabilities to accommodate estimating several statistical quantities using a {\em single} model. We can directly predict interventional probabilities, simplifying inference and enhancing outcome prediction accuracy. We demonstrate that an AR model adapted for CI is efficient and effective in various complex applications such as navigating mazes, playing chess endgames, and evaluating the impact of certain keywords on paper acceptance rates.
doi_str_mv 10.48550/arxiv.2409.18581
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2409_18581</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2409_18581</sourcerecordid><originalsourceid>FETCH-arxiv_primary_2409_185813</originalsourceid><addsrcrecordid>eNpjYJA0NNAzsTA1NdBPLKrILNMzMjGw1DO0MLUw5GSwCS3OzEtXcElNLVBwLC3JL0pNL0otLs4sS1XwzU9JzSlWSCxWcE4sLU7MUfDMS0stSs1LTlVwzUvPzEst5mFgTUvMKU7lhdLcDPJuriHOHrpge-ILijJzE4sq40H2xYPtMyasAgDmZzUH</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Using Deep Autoregressive Models as Causal Inference Engines</title><source>arXiv.org</source><creator>Im, Daniel Jiwoong ; Zhang, Kevin ; Verma, Nakul ; Cho, Kyunghyun</creator><creatorcontrib>Im, Daniel Jiwoong ; Zhang, Kevin ; Verma, Nakul ; Cho, Kyunghyun</creatorcontrib><description>Existing causal inference (CI) models are limited to primarily handling low-dimensional confounders and singleton actions. We propose an autoregressive (AR) CI framework capable of handling complex confounders and sequential actions common in modern applications. We accomplish this by {\em sequencification}, transforming data from an underlying causal diagram into a sequence of tokens. This approach not only enables training with data generated from any DAG but also extends existing CI capabilities to accommodate estimating several statistical quantities using a {\em single} model. We can directly predict interventional probabilities, simplifying inference and enhancing outcome prediction accuracy. We demonstrate that an AR model adapted for CI is efficient and effective in various complex applications such as navigating mazes, playing chess endgames, and evaluating the impact of certain keywords on paper acceptance rates.</description><identifier>DOI: 10.48550/arxiv.2409.18581</identifier><language>eng</language><subject>Computer Science - Learning ; Statistics - Machine Learning</subject><creationdate>2024-09</creationdate><rights>http://arxiv.org/licenses/nonexclusive-distrib/1.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2409.18581$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2409.18581$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Im, Daniel Jiwoong</creatorcontrib><creatorcontrib>Zhang, Kevin</creatorcontrib><creatorcontrib>Verma, Nakul</creatorcontrib><creatorcontrib>Cho, Kyunghyun</creatorcontrib><title>Using Deep Autoregressive Models as Causal Inference Engines</title><description>Existing causal inference (CI) models are limited to primarily handling low-dimensional confounders and singleton actions. We propose an autoregressive (AR) CI framework capable of handling complex confounders and sequential actions common in modern applications. We accomplish this by {\em sequencification}, transforming data from an underlying causal diagram into a sequence of tokens. This approach not only enables training with data generated from any DAG but also extends existing CI capabilities to accommodate estimating several statistical quantities using a {\em single} model. We can directly predict interventional probabilities, simplifying inference and enhancing outcome prediction accuracy. We demonstrate that an AR model adapted for CI is efficient and effective in various complex applications such as navigating mazes, playing chess endgames, and evaluating the impact of certain keywords on paper acceptance rates.</description><subject>Computer Science - Learning</subject><subject>Statistics - Machine Learning</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNpjYJA0NNAzsTA1NdBPLKrILNMzMjGw1DO0MLUw5GSwCS3OzEtXcElNLVBwLC3JL0pNL0otLs4sS1XwzU9JzSlWSCxWcE4sLU7MUfDMS0stSs1LTlVwzUvPzEst5mFgTUvMKU7lhdLcDPJuriHOHrpge-ILijJzE4sq40H2xYPtMyasAgDmZzUH</recordid><startdate>20240927</startdate><enddate>20240927</enddate><creator>Im, Daniel Jiwoong</creator><creator>Zhang, Kevin</creator><creator>Verma, Nakul</creator><creator>Cho, Kyunghyun</creator><scope>AKY</scope><scope>EPD</scope><scope>GOX</scope></search><sort><creationdate>20240927</creationdate><title>Using Deep Autoregressive Models as Causal Inference Engines</title><author>Im, Daniel Jiwoong ; Zhang, Kevin ; Verma, Nakul ; Cho, Kyunghyun</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-arxiv_primary_2409_185813</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Computer Science - Learning</topic><topic>Statistics - Machine Learning</topic><toplevel>online_resources</toplevel><creatorcontrib>Im, Daniel Jiwoong</creatorcontrib><creatorcontrib>Zhang, Kevin</creatorcontrib><creatorcontrib>Verma, Nakul</creatorcontrib><creatorcontrib>Cho, Kyunghyun</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv Statistics</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Im, Daniel Jiwoong</au><au>Zhang, Kevin</au><au>Verma, Nakul</au><au>Cho, Kyunghyun</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Using Deep Autoregressive Models as Causal Inference Engines</atitle><date>2024-09-27</date><risdate>2024</risdate><abstract>Existing causal inference (CI) models are limited to primarily handling low-dimensional confounders and singleton actions. We propose an autoregressive (AR) CI framework capable of handling complex confounders and sequential actions common in modern applications. We accomplish this by {\em sequencification}, transforming data from an underlying causal diagram into a sequence of tokens. This approach not only enables training with data generated from any DAG but also extends existing CI capabilities to accommodate estimating several statistical quantities using a {\em single} model. We can directly predict interventional probabilities, simplifying inference and enhancing outcome prediction accuracy. We demonstrate that an AR model adapted for CI is efficient and effective in various complex applications such as navigating mazes, playing chess endgames, and evaluating the impact of certain keywords on paper acceptance rates.</abstract><doi>10.48550/arxiv.2409.18581</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2409.18581
ispartof
issn
language eng
recordid cdi_arxiv_primary_2409_18581
source arXiv.org
subjects Computer Science - Learning
Statistics - Machine Learning
title Using Deep Autoregressive Models as Causal Inference Engines
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-09T04%3A24%3A36IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Using%20Deep%20Autoregressive%20Models%20as%20Causal%20Inference%20Engines&rft.au=Im,%20Daniel%20Jiwoong&rft.date=2024-09-27&rft_id=info:doi/10.48550/arxiv.2409.18581&rft_dat=%3Carxiv_GOX%3E2409_18581%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true