One-step ahead sequential Super Learning from short times series of many slightly dependent data, and anticipating the cost of natural disasters - Université Paris Cité Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2021

One-step ahead sequential Super Learning from short times series of many slightly dependent data, and anticipating the cost of natural disasters

Geoffrey Ecoto
  • Fonction : Auteur
  • PersonId : 1106431
Aurélien F Bibaut
  • Fonction : Auteur
Antoine Chambaz
  • Fonction : Auteur
  • PersonId : 867345

Résumé

Suppose that we observe a short time series where each time-t-specific data-structure consists of many slightly dependent data indexed by a and that we want to estimate a feature of the law of the experiment that depends neither on t nor on a. We develop and study an algorithm to learn sequentially which base algorithm in a user-supplied collection best carries out the estimation task in terms of excess risk and oracular inequalities. The analysis, which uses dependency graph to model the amount of conditional independence within each t-specific data-structure and a concentration inequality by Janson [2004], leverages a large ratio of the number of distinct a's to the degree of the dependency graph in the face of a small number of t-specific data-structures. The so-called one-step ahead Super Learner is applied to the motivating example where the challenge is to anticipate the cost of natural disasters in France.
Fichier principal
Vignette du fichier
osasl_v0.pdf (498.53 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03300559 , version 1 (28-07-2021)

Identifiants

Citer

Geoffrey Ecoto, Aurélien F Bibaut, Antoine Chambaz. One-step ahead sequential Super Learning from short times series of many slightly dependent data, and anticipating the cost of natural disasters. 2021. ⟨hal-03300559⟩
126 Consultations
94 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More