Domaines de Recherche: ?  Statistiques/Applications
 Informatique/Recherche opérationnelle
 Mathématiques/Statistiques
 Statistiques/Théorie

Productions scientifiques :


Simulating new multicriteria data from a given data set
Auteur(s): Rolland A., Cugliari J.
Conference: DA2PL 2016 (Paderborn, DE, 20161107)
Actes de conférence: , vol. p. ()
Résumé: Several methods have been proposed in the past decades
to deal with Multicriteria Decision Aiding (MCDA) problems. Even
if the axiomatic foundations of these methods are generally wellknown,
comparing the different methods or simply analysing the results
produced by the methods on reallife problems is arduous as
there is a lack of benchmark MCDA datasets in the literature. A
solution is therefore to simulate new MCDA dataset examples. But
the analysis of realworld examples show that one must deal with
data that are lightly linked, and then it is important to take into account
dependency between variables when simulating new datasets.
We propose in this paper three different approaches to simulate new
data based on existing small datasets. We describe these methods,
we propose a quality analysis of the results, and we experiment the
methods on different examples from the literature.



Gametheoretically Optimal Reconciliation of Contemporaneous Hierarchical Time Series Forecasts
Auteur(s): Van Erven Tim, Cugliari J.
(Document sans référence bibliographique)
Ref HAL: hal00920559_v1
Résumé: In hierarchical time series (HTS) forecasting, the hierarchical relation be tween multiple time series is exploited to make better forecasts. This hierarchical relation implies one or more aggregate consistency constraints that the series are known to satisfy. Many existing approaches, like for example bottomup or top down forecasting, therefore attempt to achieve this goal in a way that guarantees that the forecasts will also be aggregate consistent. We propose to split the problem of HTS into two independent steps: first one comes up with the best possible fore casts for the time series without worrying about aggregate consistency; and then a reconciliation procedure is used to make the forecasts aggregate consistent. We introduce a GameTheoretically OPtimal (GTOP) reconciliation method, which is guaranteed to only improve any given set of forecasts. This opens up new possibil ities for constructing the forecasts. For example, it is not necessary to assume that bottomlevel forecasts are unbiased, and aggregate forecasts may be constructed by regressing both on bottomlevel forecasts and on other covariates that may only be available at the aggregate level. We illustrate the benefits of our approach both on simulated data and on real electricity consumption data.



On the use of copulas to simulate multicriteria data
Auteur(s): Rolland A., Cugliari J., Tran Thiminhthui
Conference: DA2PL 2014 (Chatenay Malabry, FR, 20141117)
Actes de conférence: , vol. p. ()
Résumé: On the use of copulas to simulate multicriteria data



A prediction interval for a functionvalued forecast model
Auteur(s): Antoniadis Anestis, Brossat Xavier, Cugliari J., Poggi JeanMichel
(Document sans référence bibliographique) 20140801
Ref HAL: hal01094797_v1
Ref Arxiv: 1412.4222
Ref. & Cit.: NASA ADS
Résumé: Starting from the information contained in the shape of the load curves, we have proposed a flexible nonparametric functionvalued forecast model called KWF (Kernel+Wavelet+Functional) well suited to handle nonstationary series. The predictor can be seen as a weighted average of futures of past situations, where the weights increase with the similarity between the past situations and the actual one. In addition, this strategy provides with a simultaneous multiple horizon prediction. These weights induce a probability distribution that can be used to produce bootstrap pseudo predictions. Prediction intervals are constructed after obtaining the corresponding bootstrap pseudo prediction residuals. We develop two propositions following directly the KWF strategy and compare it to two alternative ways coming from proposals of econometricians. They construct simultaneous prediction intervals using multiple comparison corrections through the control of the family wise error (FWE) or the false discovery rate. Alternatively, such prediction intervals can be constructed bootstrapping joint probability regions. In this work we propose to obtain prediction intervals for the KWF model that are simultaneously valid for the H prediction horizons that corresponds with the corresponding path forecast, making a connection between functional time series and the econometricians' framework.



Making Regional Forecast add up
Auteur(s): Van Erven Tim, Cugliari J.
(Article) Publié:
Lecture Notes In Statistics Modeling And Stochastic Learning For Forecasting In High Dimension, vol. p. (2014)
