ODE Discovery for Longitudinal Heterogeneous Treatment Effects Inference
Inferring unbiased treatment effects has received widespread attention in the machine learning community. In recent years, our community has proposed numerous solutions in standard settings, high-dimensional treatment settings, and even longitudinal settings. While very diverse, the solution has mos...
Saved in:
Main Authors | , , , , |
---|---|
Format | Journal Article |
Language | English |
Published |
15.03.2024
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Abstract | Inferring unbiased treatment effects has received widespread attention in the
machine learning community. In recent years, our community has proposed
numerous solutions in standard settings, high-dimensional treatment settings,
and even longitudinal settings. While very diverse, the solution has mostly
relied on neural networks for inference and simultaneous correction of
assignment bias. New approaches typically build on top of previous approaches
by proposing new (or refined) architectures and learning algorithms. However,
the end result -- a neural-network-based inference machine -- remains
unchallenged. In this paper, we introduce a different type of solution in the
longitudinal setting: a closed-form ordinary differential equation (ODE). While
we still rely on continuous optimization to learn an ODE, the resulting
inference machine is no longer a neural network. Doing so yields several
advantages such as interpretability, irregular sampling, and a different set of
identification assumptions. Above all, we consider the introduction of a
completely new type of solution to be our most important contribution as it may
spark entirely new innovations in treatment effects in general. We facilitate
this by formulating our contribution as a framework that can transform any ODE
discovery method into a treatment effects method. |
---|---|
AbstractList | Inferring unbiased treatment effects has received widespread attention in the
machine learning community. In recent years, our community has proposed
numerous solutions in standard settings, high-dimensional treatment settings,
and even longitudinal settings. While very diverse, the solution has mostly
relied on neural networks for inference and simultaneous correction of
assignment bias. New approaches typically build on top of previous approaches
by proposing new (or refined) architectures and learning algorithms. However,
the end result -- a neural-network-based inference machine -- remains
unchallenged. In this paper, we introduce a different type of solution in the
longitudinal setting: a closed-form ordinary differential equation (ODE). While
we still rely on continuous optimization to learn an ODE, the resulting
inference machine is no longer a neural network. Doing so yields several
advantages such as interpretability, irregular sampling, and a different set of
identification assumptions. Above all, we consider the introduction of a
completely new type of solution to be our most important contribution as it may
spark entirely new innovations in treatment effects in general. We facilitate
this by formulating our contribution as a framework that can transform any ODE
discovery method into a treatment effects method. |
Author | Qian, Zhaozhi Berrevoets, Jeroen Kacprzyk, Krzysztof Holt, Samuel van der Schaar, Mihaela |
Author_xml | – sequence: 1 givenname: Krzysztof surname: Kacprzyk fullname: Kacprzyk, Krzysztof – sequence: 2 givenname: Samuel surname: Holt fullname: Holt, Samuel – sequence: 3 givenname: Jeroen surname: Berrevoets fullname: Berrevoets, Jeroen – sequence: 4 givenname: Zhaozhi surname: Qian fullname: Qian, Zhaozhi – sequence: 5 givenname: Mihaela surname: van der Schaar fullname: van der Schaar, Mihaela |
BackLink | https://doi.org/10.48550/arXiv.2403.10766$$DView paper in arXiv |
BookMark | eNotz71OwzAUQGEPMJTCAzDVL5AQx78ZURtIpUhdskfXyXVlqbWR41b07RGF6WxH-p7IQ4gBCXllVSmMlNUbpG9_LWtR8ZJVWqkV6Q67lu78MsUrpht1MdE-hqPPl9kHONEOM6Z4xIDxstAhIeQzhkxb53DKC90HhwnDhM_k0cFpwZf_rsnw0Q7brugPn_vte1-A0qpA2TijhRLAGTPCVbaepHFKaV7PVs0CpbBGc45mhoY5hdI0WltmQVtAwddk87e9U8av5M-QbuMvabyT-A8_ikh7 |
ContentType | Journal Article |
Copyright | http://creativecommons.org/licenses/by/4.0 |
Copyright_xml | – notice: http://creativecommons.org/licenses/by/4.0 |
DBID | AKY EPD GOX |
DOI | 10.48550/arxiv.2403.10766 |
DatabaseName | arXiv Computer Science arXiv Statistics arXiv.org |
DatabaseTitleList | |
Database_xml | – sequence: 1 dbid: GOX name: arXiv.org url: http://arxiv.org/find sourceTypes: Open Access Repository |
DeliveryMethod | fulltext_linktorsrc |
ExternalDocumentID | 2403_10766 |
GroupedDBID | AKY EPD GOX |
ID | FETCH-LOGICAL-a676-e59f87464a31184f0b2c58f66732db6d4e54b8733e8da91f6e58977b1ba7bae43 |
IEDL.DBID | GOX |
IngestDate | Wed Mar 20 12:39:02 EDT 2024 |
IsDoiOpenAccess | true |
IsOpenAccess | true |
IsPeerReviewed | false |
IsScholarly | false |
Language | English |
LinkModel | DirectLink |
MergedId | FETCHMERGED-LOGICAL-a676-e59f87464a31184f0b2c58f66732db6d4e54b8733e8da91f6e58977b1ba7bae43 |
OpenAccessLink | https://arxiv.org/abs/2403.10766 |
ParticipantIDs | arxiv_primary_2403_10766 |
PublicationCentury | 2000 |
PublicationDate | 2024-03-15 |
PublicationDateYYYYMMDD | 2024-03-15 |
PublicationDate_xml | – month: 03 year: 2024 text: 2024-03-15 day: 15 |
PublicationDecade | 2020 |
PublicationYear | 2024 |
Score | 1.9120663 |
SecondaryResourceType | preprint |
Snippet | Inferring unbiased treatment effects has received widespread attention in the
machine learning community. In recent years, our community has proposed
numerous... |
SourceID | arxiv |
SourceType | Open Access Repository |
SubjectTerms | Computer Science - Learning Statistics - Methodology |
Title | ODE Discovery for Longitudinal Heterogeneous Treatment Effects Inference |
URI | https://arxiv.org/abs/2403.10766 |
hasFullText | 1 |
inHoldings | 1 |
isFullTextHit | |
isPrint | |
link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwdV09T8MwELXaTiwIBKh8ygOrRZz4c0S0JSCgS5C6Vb74LHVpUT8Q_HvsJBUsrPYtvhvunfzuPUJuTcg0eAjMIhdMeA7MAkqW1hqtDZnXjTr_65sq38XzTM56hO53Ydz6a_HZ6gPD5i6JxcX5UivVJ_08T5Stx-ms_ZxspLi6-N-4iDGboz9NYnJEDjt0R-_bchyTHi5PSDkdjelosakTW_KbRpRIX1bJJmjnkyUVLRMjZRULiXEKp9We-U1bYeENfdrv5J2SajKuHkrWGRgwp7RiKG0wWijhigjjRcggr6UJyWgz96C8QCnA6KJA453lQaE0EY4BB6fBoSjOyGC5WuIwEZAsGomh4GgE-hpUDdxi4E6EwtnsnAybZ88_Wo2KecrIvMnIxf9Xl-Qgjz06Uaq4vCKD7XqH17HHbuGmSfQPdyB8-g |
link.rule.ids | 228,230,786,891 |
linkProvider | Cornell University |
openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=ODE+Discovery+for+Longitudinal+Heterogeneous+Treatment+Effects+Inference&rft.au=Kacprzyk%2C+Krzysztof&rft.au=Holt%2C+Samuel&rft.au=Berrevoets%2C+Jeroen&rft.au=Qian%2C+Zhaozhi&rft.date=2024-03-15&rft_id=info:doi/10.48550%2Farxiv.2403.10766&rft.externalDocID=2403_10766 |