Communication-efficient Conformal Prediction for Distributed Datasets

Nery Riquelme-Granada; Zhiyuan Luo; Khuong An Nguyen

Communication-efficient Conformal Prediction for Distributed Datasets

Nery Riquelme-Granada, Zhiyuan Luo, Khuong An Nguyen

Research output: Contribution to conference › Paper › peer-review

Abstract

Coresets have been proven useful in accelerating the computation of inductive conformal predictors (ICP) when the training data becomes large in size.This work shows that coreset-based conformal predictors are not only computationally efficient in the centralised setting, but may also naturally be used in scenarios where the dataset of interested in inherently distributed.

Original language	English
Pages	310-312
Number of pages	3
Publication status	Published - 26 Aug 2022

Access to Document

Cite this

@conference{5f6d348178fc46568a05e0ad8acb60e6,

title = "Communication-efficient Conformal Prediction for Distributed Datasets",

abstract = "Coresets have been proven useful in accelerating the computation of inductive conformal predictors (ICP) when the training data becomes large in size.This work shows that coreset-based conformal predictors are not only computationally efficient in the centralised setting, but may also naturally be used in scenarios where the dataset of interested in inherently distributed.",

author = "Nery Riquelme-Granada and Zhiyuan Luo and Nguyen, {Khuong An}",

note = "11th Symposium on Conformal and Probabilistic Prediction with Applications, COPA 2022 ; Conference date: 24-08-2022 Through 26-08-2022",

year = "2022",

month = aug,

day = "26",

language = "English",

pages = "310--312",

}

TY - CONF

T1 - Communication-efficient Conformal Prediction for Distributed Datasets

AU - Riquelme-Granada, Nery

AU - Luo, Zhiyuan

AU - Nguyen, Khuong An

N1 - 11th Symposium on Conformal and Probabilistic Prediction with Applications, COPA 2022 ; Conference date: 24-08-2022 Through 26-08-2022

PY - 2022/8/26

Y1 - 2022/8/26

N2 - Coresets have been proven useful in accelerating the computation of inductive conformal predictors (ICP) when the training data becomes large in size.This work shows that coreset-based conformal predictors are not only computationally efficient in the centralised setting, but may also naturally be used in scenarios where the dataset of interested in inherently distributed.

AB - Coresets have been proven useful in accelerating the computation of inductive conformal predictors (ICP) when the training data becomes large in size.This work shows that coreset-based conformal predictors are not only computationally efficient in the centralised setting, but may also naturally be used in scenarios where the dataset of interested in inherently distributed.

M3 - Paper

SP - 310

EP - 312

ER -