Université de Moncton

Communications, affaires publiques et marketing
Fil de nouvelles Bookmark and Share
Mardi 22 Janvier 2013

Jules de Tibeiro publie deux articles dans le Journal of Classification

Jules de Tibeiro, professeur à l'UMCS.
Agrandir l'image
Le professeur Jules de Tibeiro vient de publier, avec ses collègues, deux nouveaux articles dans le journal (arbitré) Journal of Classification (Classification Society of North
America with the support of the International Federation of Classification Societies) par
Springer-Verlag New York, Inc. Ces articles sont intitulés :

1. “Graph Partitioning by Correspondence Analysis and
Taxicab Correspondence Analysis”
. CC: joc@fsw.leidenuniv.nl
30 (2013) DOI: 10.1007/s00357-
Auteurs : V. CHOULAKIAN* and J. J. S. de TIBEIRO**
*Université de Moncton, campus de Moncton, N.-B.
**Université de Moncton, campus de Shippagan, N.-B.
Abstract: We consider Correspondence Analysis (CA) and Taxicab Correspondence Analysis (TCA) of relational datasets that can mathematically be described as weighted loopless graphs. Such data appears in particular in Network Analysis. We present CA and TCA as relaxation methods for the graph partitioning problem. Examples of real datasets are provided.
Key words: Adjacency matrix; Incidence matrix; Network analysis; Graph partitioning; Graph Laplacian matrix; Correspondence analysis; Taxicab correspondence analysis, NCut; RCut; MCut; Matrix norm; Centroid method.

2. “Optimal Quantization of the Support of a Continuous Multivariate Distribution Based on Mutual Information”. CC: joc@fsw.leidenuniv.nl
CLAS213R1
Auteurs : B. COLIN*, F. DUBEAU*, H. KHREIBANI* and Jules J.S. de TIBEIRO**
*Université de Sherbrooke, Qc
**Université de Moncton, campus de Shippagan, N.-B.
Abstract: Based on the notion of Mutual Information between the components of a
random vector, we construct, for data reduction reasons, an optimal quantization of the support of its probability measure. More precisely, we propose a simultaneous discretization of the whole set of the components of the random vector which takes into account, as much as possible, the stochastic dependence between them. Examples are presented.
Source : Service des communications, Campus de Shippagan
Communiqué de presse - Pour diffusion immédiate -
Fil de nouvelles Bookmark and Share
Imprimer cette page
Calendrier
Nouvelles
Bottin
Accueil  |  Communications, affaires publiques et marketing  |  Urgence  |  Assistance technique © 2017, Université de Moncton. Tous droits réservés.