AdaptiveConformal: An R Package for Adaptive Conformal Inference

AdaptiveConformal: An R Package for Adaptive Conformal Inference

Creative Commons BY License ISSN 2824-7795

Conformal Inference (CI) is a popular approach for generating finite sample prediction intervals based on the output of any point prediction method when data are exchangeable. Adaptive Conformal Inference (ACI) algorithms extend CI to the case of sequentially observed data, such as time series, and exhibit strong theoretical guarantees without having to assume exchangeability of the observed data. The common thread that unites algorithms in the ACI family is that they adaptively adjust the width of the generated prediction intervals in response to the observed data. We provide a detailed description of five ACI algorithms and their theoretical guarantees, and test their performance in simulation studies. We then present a case study of producing prediction intervals for influenza incidence in the United States based on black-box point forecasts. Implementations of all the algorithms are released as an open-source R package, AdaptiveConformal, which also includes tools for visualizing and summarizing conformal prediction intervals.

Authors
Affiliations
Published

July 18, 2024

Modified

August 9, 2024

Keywords

Conformal inference, Adaptive conformal inference, time series, R

Status
build status

reviews
Abstract

Conformal Inference (CI) is a popular approach for generating finite sample prediction intervals based on the output of any point prediction method when data are exchangeable. Adaptive Conformal Inference (ACI) algorithms extend CI to the case of sequentially observed data, such as time series, and exhibit strong theoretical guarantees without having to assume exchangeability of the observed data. The common thread that unites algorithms in the ACI family is that they adaptively adjust the width of the generated prediction intervals in response to the observed data. We provide a detailed description of five ACI algorithms and their theoretical guarantees, and test their performance in simulation studies. We then present a case study of producing prediction intervals for influenza incidence in the United States based on black-box point forecasts. Implementations of all the algorithms are released as an open-source R package, AdaptiveConformal, which also includes tools for visualizing and summarizing conformal prediction intervals.

1 Introduction

Ce document est un exemple de fichier Quarto Markdown (.qmd) qui utilise Python, Jupyter, nbformat et numpy.

2 Utilisation de Python

Hide/Show the code
print("Bonjour, Quarto!")
Bonjour, Quarto!
Hide/Show the code
import nbformat
from nbformat import read

# Exemple de lecture d'un notebook Jupyter
notebook_path = 'notebook.ipynb' 
with open(notebook_path, 'r', encoding='utf-8') as f:
    nb = read(f, as_version=4)

# Afficher les métadonnées du notebook
print(nb.metadata)
{'language_info': {'name': 'python'}}
Hide/Show the code
import numpy as np

# Création d'un tableau numpy
a = np.array([1, 2, 3, 4, 5])
print("Tableau numpy:", a)

# Calcul de la moyenne
moyenne = np.mean(a)
print("Moyenne:", moyenne)

# Calcul de l'écart type
ecart_type = np.std(a)
print("Écart type:", ecart_type)
Tableau numpy: [1 2 3 4 5]
Moyenne: 3.0
Écart type: 1.4142135623730951
Back to top

Reuse

Citation

BibTeX citation:
@article{susmann2024,
  author = {Susmann, Herbert and Chambaz, Antoine and Josse, Julie},
  publisher = {Société Française de Statistique},
  title = {AdaptiveConformal: {An} {`R`} {Package} for {Adaptive}
    {Conformal} {Inference}},
  journal = {Computo},
  date = {2024-07-18},
  url = {https://computo.sfds.asso.fr/template-computo-quarto},
  doi = {10.57750/edan-5f53},
  issn = {2824-7795},
  langid = {en},
  abstract = {Conformal Inference (CI) is a popular approach for
    generating finite sample prediction intervals based on the output of
    any point prediction method when data are exchangeable. Adaptive
    Conformal Inference (ACI) algorithms extend CI to the case of
    sequentially observed data, such as time series, and exhibit strong
    theoretical guarantees without having to assume exchangeability of
    the observed data. The common thread that unites algorithms in the
    ACI family is that they adaptively adjust the width of the generated
    prediction intervals in response to the observed data. We provide a
    detailed description of five ACI algorithms and their theoretical
    guarantees, and test their performance in simulation studies. We
    then present a case study of producing prediction intervals for
    influenza incidence in the United States based on black-box point
    forecasts. Implementations of all the algorithms are released as an
    open-source `R` package, `AdaptiveConformal`, which also includes
    tools for visualizing and summarizing conformal prediction
    intervals.}
}
For attribution, please cite this work as:
Susmann, Herbert, Antoine Chambaz, and Julie Josse. 2024. “AdaptiveConformal: An `R` Package for Adaptive Conformal Inference.” Computo, July. https://doi.org/10.57750/edan-5f53.