2022
DOI: 10.48550/arxiv.2203.01761
|View full text |Cite
Preprint
|
Sign up to set email alerts
|

Doubly Robust Calibration of Prediction Sets under Covariate Shift

Abstract: Conformal prediction has received tremendous attention in recent years with several applications both in health and social sciences. Recently, conformal inference has offered new solutions to problems in causal inference, which have leveraged advances in modern discipline of semiparametric statistics to construct efficient tools for prediction uncertainty quantification. In this paper, we consider the problem of obtaining distribution-free prediction regions accounting for a shift in the distribution of the co… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1

Citation Types

0
3
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
2
1

Relationship

1
2

Authors

Journals

citations
Cited by 3 publications
(3 citation statements)
references
References 20 publications
0
3
0
Order By: Relevance
“…However, real-world datasets often deviate from this assumption, particularly in scenarios involving temporal dependence, such as time-series data; see Candès (2021, 2022), Zaffran et al (2022), and Bhatnagar et al (2023). In this regard, several extensions of conformal prediction techniques have addressed challenges related to distribution shift, employing methods such as reweighting and distributionally robust optimization to maintain approximately valid coverage; see Tibshirani et al (2019), Podkopaev and Ramdas (2021), Yang et al (2022), and Barber et al (2023).…”
Section: Literature Reviewmentioning
confidence: 99%
“…However, real-world datasets often deviate from this assumption, particularly in scenarios involving temporal dependence, such as time-series data; see Candès (2021, 2022), Zaffran et al (2022), and Bhatnagar et al (2023). In this regard, several extensions of conformal prediction techniques have addressed challenges related to distribution shift, employing methods such as reweighting and distributionally robust optimization to maintain approximately valid coverage; see Tibshirani et al (2019), Podkopaev and Ramdas (2021), Yang et al (2022), and Barber et al (2023).…”
Section: Literature Reviewmentioning
confidence: 99%
“…However, real-world datasets often deviate from this assumption, particularly in scenarios involving temporal dependence, such as time series data, see [Gibbs and Candes, 2021], [Zaffran et al, 2022], [Gibbs and Candès, 2022], and [Bhatnagar et al, 2023]. In this regard, several extensions of conformal prediction techniques have addressed challenges related to distribution shift, employing methods such as reweighting and distributionally robust optimization to maintain approximately valid coverage, see [Tibshirani et al, 2019], [Podkopaev and Ramdas, 2021], [Yang et al, 2022], and [Barber et al, 2023].…”
Section: Literature Reviewmentioning
confidence: 99%
“…where Pr(Y / ∈ C(X) | A = 0) is an estimator of Pr P 0 (Y / ∈ C(X) | A = 0) and size(C) is measure of the size of C. This approach has been considered in Yang et al [96], and generally results in smaller prediction sets than the APAC ones. The reason is that the APAC guarantee requires approximately controlling the confidence level 1 − α conf over the data at hand, which leads to some conservativeness.…”
Section: Basic Settingmentioning
confidence: 99%