seminari INRAE

Seminari di Jean-Michel Roger e Maxime Ryckewaert

Giovedì, 4 Novembre, 2021

Federico Marini è lieto di invitarvi ai seminari di dipartimento:

N-CovSel, a new strategy for feature selection in N-way data

di Jean-Michel Roger
INRAE's Occitanie-Montpellier Centre (France)


In data analysis, how to select meaningful variables is a hot and wide-debated topic and several variable selection (or feature reduction) approaches have been proposed into the literature.

Covariance Selection (CovSel) is an existing method which is conceived to select variables in regression and discrimination contexts, and it assesses the features’ relevancy based on their covariance with the response(s). Most of variable selection methods refer to contexts in which data are collected in matrices. How to assess the relevancy of variables in a multi- way context has not been extensively discussed yet.

The present contribution, named N-CovSel, proposes to extend the CovSel principle to the N-Way structures, by selecting features in place of variables. Three main questions are addressed to achieve this: (i) How to define a feature in a N-Way array (Figure 1); (ii) How to define the covariance between a feature and a response Y; (iii) How to deflate a N-Way array with regard to a selected feature.

The complete algorithm of N-CovSel will be presented and its theoretical properties discussed. Two applications on 3-way real data will be presented, illustrating that the proposed method can be differently used, depending on the final purpose of the analysis.

Reduction of repeatability error for analysis of variance-Simultaneous Component Analysis (REP-ASCA): Application to NIR spectroscopy

di Maxime Ryckewaert
INRAE's Occitanie-Montpellier Centre (France)


For multivariate data associated with an experimental design, Analysis of Variance - Simultaneous Component Analysis (ASCA) is the most commonly used. However, when faced with measurements that lack repeatability, results of ASCA can then be misinterpreted.

A method named Reduction of the error of repeatability-Analysis of variance-Simultaneous

Component Analysis (REP-ASCA) has been developed to address this issue. This approach proposes to adapt a protocol consisting of adding repeated measures to the measurements for the analysis of variance. The methodology of REP-ASCA will be detailed and applied to two case studies: NIR spectroscopy of coffee data and VIS-NIR spectroscopy for maize for plant breeding.

I seminari si terranno in Sala Parravano giovedì 4 novembre dalle ore 15.00.

Potranno essere seguiti in presenza (il numero dei posti disponibili in Sala Parravano è limitato a 32 dalle misure di prevenzione e protezione dal contagio da COVID-19 indicate nei DPCM del 26/04/2020 e successivi e alle misure aggiuntive indicate dall'Ateneo.

È possibile partecipare da remoto su piattaforma Google Meet:


© Università degli Studi di Roma "La Sapienza" - Piazzale Aldo Moro 5, 00185 Roma