{"id":16745,"date":"2021-06-27T13:09:33","date_gmt":"2021-06-27T11:09:33","guid":{"rendered":"https:\/\/www.esi-sba.dz\/fr\/?post_type=tribe_events&p=16745"},"modified":"2021-06-27T17:27:37","modified_gmt":"2021-06-27T15:27:37","slug":"soutenance-de-doctorat-attaoui-mohammed-oualid","status":"publish","type":"tribe_events","link":"https:\/\/www.esi-sba.dz\/fr\/index.php\/evenements-calendar\/soutenance-de-doctorat-attaoui-mohammed-oualid\/","title":{"rendered":"Soutenance de Doctorat : ATTAOUI Mohammed Oualid"},"content":{"rendered":"

\"\"<\/p>\n

Nom et Pr\u00e9nom du Doctorant<\/strong> : ATTAOUI Mohammed Oualid
\nSp\u00e9cialit\u00e9<\/strong> : Syst\u00e8me d\u2019Information
\nIntitul\u00e9 de la th\u00e8se<\/strong> : Vers de nouvelles m\u00e9thodes de clustering de flux de donn\u00e9es.
\nDate et lieu<\/strong> : aura lieu (\u00e0 huis clos en raison des mesures de restrictions sanitaires) le Mercredi 30 Juin 2021 \u00e0 10H00 \u00e0 la salle de soutenance de l\u2019\u00e9cole.<\/strong><\/p>\n

Soutenance de th\u00e8se de doctorat en cotutelle avec l\u2019universit\u00e9 Sorbonne Paris Nord, France devant le jury :<\/p>\n\n\n\n\n\n\n\n\n\n\n\n
Nom et pr\u00e9noms<\/strong><\/td>\nGrade<\/strong><\/td>\n\u00c9tablissement d\u2019appartenance<\/strong><\/td>\nQualit\u00e9<\/strong><\/td>\n<\/tr>\n
BENSLIMANE Sidi Mohamed<\/td>\nPr<\/td>\nESI de Sidi Bel Abbes<\/td>\nPr\u00e9sident<\/td>\n<\/tr>\n
KESKES Nabil<\/td>\nPr.<\/td>\nESI de Sidi Bel Abbes<\/td>\nDirecteur de th\u00e8se<\/td>\n<\/tr>\n
LEBBAH Mustapha<\/td>\nMC HDR<\/td>\nUniversit\u00e9 Sorbonne Paris<\/p>\n

Nord, France<\/td>\n

Directeur de th\u00e8se<\/td>\n<\/tr>\n
BOUCHAFFRA Djamel<\/td>\nDirecteur de recherches<\/td>\nCDTA, Alger<\/td>\nExaminateur<\/td>\n<\/tr>\n
CHARNOIS Thierry<\/td>\nPr.<\/td>\nSorbonne Paris<\/p>\n

Nord, France<\/td>\n

Examinateur<\/td>\n<\/tr>\n
DE RUNZ Cyril<\/td>\nMC HDR<\/td>\nUniversit\u00e9 de Tours<\/td>\nExaminateur<\/td>\n<\/tr>\n
FORESTIER Germain<\/td>\nPr.<\/td>\nUniversit\u00e9 d\u2019Haute-Alsace, France<\/td>\nInvit\u00e9<\/td>\n<\/tr>\n
AZZAG Hanene<\/td>\nMC HDR<\/td>\nUniversit\u00e9 Sorbonne Paris, France<\/td>\nInvit\u00e9e<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

 <\/p>\n

[R\u00e9sum\u00e9:]
\nCes derniers jours, de grandes quantit\u00e9s de donn\u00e9es sont g\u00e9n\u00e9r\u00e9es par les applications en temps r\u00e9el. Ces quantit\u00e9s de donn\u00e9es appel\u00e9es flux de donn\u00e9es ne peuvent pas \u00eatre trait\u00e9es comme des donn\u00e9es classiques car nous ne pouvons pas stocker ou traiter cette quantit\u00e9 de donn\u00e9es. L’exploration de flux est le processus qui consiste \u00e0 trouver une structure complexe dans un grand volume de donn\u00e9es o\u00f9 les donn\u00e9es \u00e9voluent et arrivent dans un flux non limit\u00e9. Un flux de donn\u00e9es est une s\u00e9quence de donn\u00e9es continues qui impose une restriction de passage unique. L’acc\u00e8s al\u00e9atoire aux donn\u00e9es n’est pas possible, et il est peu pratique de stocker toutes les donn\u00e9es qui arrivent. Dans ce cas, nous stockons des caract\u00e9ristiques ou des synopsis de clusters qui comprennent g\u00e9n\u00e9ralement des statistiques descriptives pour un cluster. Dans de nombreux cas, les algorithmes de flux de donn\u00e9es doivent respecter des contraintes d’espace et de temps. Les donn\u00e9es arrivant dans les flux contiennent souvent du bruit et des valeurs aberrantes. Ainsi, le clustering de flux de donn\u00e9es doit d\u00e9tecter, distinguer et filtrer ces donn\u00e9es avant la t\u00e2che de clustering. Le pr\u00e9sent travail porte sur la mod\u00e9lisation de donn\u00e9es \u00e0 haute dimension dans un cadre de flux de donn\u00e9es, en utilisant le Subspace Clustering pour d\u00e9couvrir des clusters int\u00e9gr\u00e9s dans diff\u00e9rents sous-espaces. Nous avons \u00e9galement utilis\u00e9 le cadre multi-objectif pour faire face aux variations des caract\u00e9ristiques des donn\u00e9es. Nous avons pr\u00e9sent\u00e9 diff\u00e9rentes techniques bas\u00e9es sur le subspace clustering, le Multi-Objective clustering combin\u00e9 avec des techniques d’analyse de flux pour r\u00e9pondre aux probl\u00e8mes mentionn\u00e9s pr\u00e9c\u00e9demment.<\/p>\n

[Abstract:]
\nThese days, large amounts of data are generated by real-time applications. These amounts of data called data streams cannot be processed as conventional data because we cannot store or process this amount of data. Stream mining is the process of finding a complex structure in a large volume of data where data evolves and arrives in an unbounded stream. A data stream is a sequence of continuous data that imposes a single passage restriction. Random access to data is not possible, and it is impractical to store all the data that arrives. In this case, we store features or synopses of clusters that typically include descriptive statistics for a cluster. In many cases, data stream algorithms must meet memory and time constraints. Data arriving in streams often contain noise and outliers. Thus, data stream clustering must detect, distinguish, and filter these data prior to the clustering task. The present work focuses on modeling high-dimensional data in a data stream framework, using Subspace Clustering to discover embedded clusters in different subspaces. We also used the multi-objective framework to cope with variations in data characteristics.\u00a0 We have presented different techniques based on Subspace Clustering, Multi-Objective clustering combined with stream analysis techniques to address the above mentioned problems.<\/p>\n

 <\/p>\n","protected":false},"excerpt":{"rendered":"

Nom et Pr\u00e9nom du Doctorant : ATTAOUI Mohammed Oualid Sp\u00e9cialit\u00e9 : Syst\u00e8me d\u2019Information Intitul\u00e9 de la th\u00e8se : Vers de nouvelles m\u00e9thodes de clustering de flux de donn\u00e9es. Date et […]<\/p>\n","protected":false},"author":1,"featured_media":16751,"template":"","meta":{"_eb_attr":"","_tribe_events_status":"","_tribe_events_status_reason":"","footnotes":""},"tags":[],"tribe_events_cat":[87],"class_list":["post-16745","tribe_events","type-tribe_events","status-publish","has-post-thumbnail","hentry","tribe_events_cat-tous","cat_tous"],"_links":{"self":[{"href":"https:\/\/www.esi-sba.dz\/fr\/index.php\/wp-json\/wp\/v2\/tribe_events\/16745","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.esi-sba.dz\/fr\/index.php\/wp-json\/wp\/v2\/tribe_events"}],"about":[{"href":"https:\/\/www.esi-sba.dz\/fr\/index.php\/wp-json\/wp\/v2\/types\/tribe_events"}],"author":[{"embeddable":true,"href":"https:\/\/www.esi-sba.dz\/fr\/index.php\/wp-json\/wp\/v2\/users\/1"}],"version-history":[{"count":6,"href":"https:\/\/www.esi-sba.dz\/fr\/index.php\/wp-json\/wp\/v2\/tribe_events\/16745\/revisions"}],"predecessor-version":[{"id":16779,"href":"https:\/\/www.esi-sba.dz\/fr\/index.php\/wp-json\/wp\/v2\/tribe_events\/16745\/revisions\/16779"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.esi-sba.dz\/fr\/index.php\/wp-json\/wp\/v2\/media\/16751"}],"wp:attachment":[{"href":"https:\/\/www.esi-sba.dz\/fr\/index.php\/wp-json\/wp\/v2\/media?parent=16745"}],"wp:term":[{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.esi-sba.dz\/fr\/index.php\/wp-json\/wp\/v2\/tags?post=16745"},{"taxonomy":"tribe_events_cat","embeddable":true,"href":"https:\/\/www.esi-sba.dz\/fr\/index.php\/wp-json\/wp\/v2\/tribe_events_cat?post=16745"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}