國立高雄大學圖資館 |

登入

回首頁

Statistical Methods for Causal Inference With Observational Data: Heterogeneity and Uncertain Exposures /

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Statistical Methods for Causal Inference With Observational Data: Heterogeneity and Uncertain Exposures /Ajinkya Hemant Kokandakar.
作者:	Kokandakar, Ajinkya Hemant,
面頁冊數:	1 electronic resource (136 pages)
附註:	Source: Dissertations Abstracts International, Volume: 87-02, Section: B.
附註:	Advisors: Deshpande, Sameer K. Committee members: Cochran, Amy L.; Kang, Hyunseung; Katzfuß, Matthias; Larget, Bret.
Contained By:	Dissertations Abstracts International87-02B.
標題:	Statistics.
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=32236793
ISBN:	9798290963297

Statistical Methods for Causal Inference With Observational Data: Heterogeneity and Uncertain Exposures /
Kokandakar, Ajinkya Hemant,

Statistical Methods for Causal Inference With Observational Data: Heterogeneity and Uncertain Exposures /Ajinkya Hemant Kokandakar. - 1 electronic resource (136 pages)

Source: Dissertations Abstracts International, Volume: 87-02, Section: B.

Statistical methods for causal inference are geared towards studying the effect of a specific and well-defined intervention on a well-specified outcome of interest. In this dissertation we develop statistical tools to perform causal inference when the exposure is uncertain in the sense that there can be many possible definitions of the exposure condition or the exposure is well-defined but not measured for all units. When dealing with multiple possible exposure definitions, we define a hierarchy of exposure definitions and use a hierarchical matching and testing procedure to test the statistical significance of the effect of the exposures while maintaining family wise error rate. We also develop and explore related statistical tools to better account for effect heterogeneity and uncertainty due to estimation. We use the Bayesian Causal Forests model for estimating heterogenous treatment effects in a longitudinal dataset. We implement extensions of the Varying Coefficients BART model to estimate heterogenous treatment effects and use VCBART to study the role of imputing the exposure when some units in the causal analysis do not have measurements for the exposure.

English

ISBN: 9798290963297Subjects--Topical Terms:

182057
Statistics.
Subjects--Index Terms:

Causal inference

Statistical Methods for Causal Inference With Observational Data: Heterogeneity and Uncertain Exposures /
LDR:02766nmm a22004453i 4500 001 693752
005 20260511063020.5
006 m o d
007 cr|nu||||||||
008 260819s2025 miu||||||m |||||||eng d
020 $a 9798290963297
035 $a (MiAaPQD)AAI32236793
035 $a AAI32236793
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Kokandakar, Ajinkya Hemant, $e author. $3 1010411
245 1 0 $a Statistical Methods for Causal Inference With Observational Data: Heterogeneity and Uncertain Exposures / $c Ajinkya Hemant Kokandakar.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2025
300 $a 1 electronic resource (136 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 87-02, Section: B.
500 $a Advisors: Deshpande, Sameer K. Committee members: Cochran, Amy L.; Kang, Hyunseung; Katzfuß, Matthias; Larget, Bret.
502 $b Ph.D. $c The University of Wisconsin - Madison $d 2025.
520 $a Statistical methods for causal inference are geared towards studying the effect of a specific and well-defined intervention on a well-specified outcome of interest. In this dissertation we develop statistical tools to perform causal inference when the exposure is uncertain in the sense that there can be many possible definitions of the exposure condition or the exposure is well-defined but not measured for all units. When dealing with multiple possible exposure definitions, we define a hierarchy of exposure definitions and use a hierarchical matching and testing procedure to test the statistical significance of the effect of the exposures while maintaining family wise error rate. We also develop and explore related statistical tools to better account for effect heterogeneity and uncertainty due to estimation. We use the Bayesian Causal Forests model for estimating heterogenous treatment effects in a longitudinal dataset. We implement extensions of the Varying Coefficients BART model to estimate heterogenous treatment effects and use VCBART to study the role of imputing the exposure when some units in the causal analysis do not have measurements for the exposure.
546 $a English
590 $a School code: 0262
650 4 $a Statistics. $3 182057
650 4 $a Public health. $3 209444
650 4 $a Biostatistics. $3 339693
650 4 $a Epidemiology. $3 191577
653 $a Causal inference
653 $a Observational data
653 $a Heterogeneity
653 $a Measurement error
653 $a Statistical modeling
690 $a 0463
690 $a 0766
690 $a 0308
690 $a 0573
710 2 $a The University of Wisconsin - Madison. $b Statistics. $3 857397
720 1 $a Deshpande, Sameer K. $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 87-02B.
790 $a 0262
791 $a Ph.D.
792 $a 2025
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=32236793