Recherche

Toutes nos publications

La liste com­plète des publi­ca­tions de l’Institut Henri Fayol peut être consul­tée sur l’ar­chive HAL CCSD CNRS via notre portail HAL-EMSE.

Nos dernières publications sur l’archive HAL CCSD CNRS

  • The research Chair OQUAIDO – for « Optimisation et QUAntification d’Incertitudes pour les Données Onéreuses » in French – gathers aca­de­mic and tech­no­lo­gi­cal research part­ners to work on sta­tis­ti­cal lear­ning pro­blems invol­ving scarce and error‐​prone data. This Chair, was crea­ted in January 2016 for a per­iod of 5 years. It has tack­led pro­blems where small data is […]
  • Nowadays, IoT devices have an enlar­ging scope of acti­vi­ties span­ning from sen­sing, com­pu­ting to acting and even more, lear­ning, rea­so­ning and plan­ning. As the num­ber of IoT appli­ca­tions increases, these objects are beco­ming more and more ubi­qui­tous. Therefore, they need to adapt their func­tio­na­li­ty in res­ponse to the uncer­tain­ties of their envi­ron­ment to achieve their […]
  • The desi­gn of a space trans­por­ta­tion sys­tem is a com­plex mul­ti­dis­ci­pli­na­ry opti­mi­za­tion pro­blem. Uncertainty exis­ting at the ear­ly phases of the desi­gn pro­cess influences the final sys­tem per­for­mances and relia­bi­li­ty. Uncertainty‐​based Multidisciplinary Design Optimization (UMDO) metho­do­lo­gies allow to deter­mine the opti­mal desi­gn solu­tion with res­pect to the per­for­mance cri­te­ria while ensu­ring the relia­bi­li­ty and the […]
  • The deve­lop­ment of auto­no­mous vehicles, capable of peer‐​to‐​peer com­mu­ni­ca­tion, as well as the inter­est in on‐​demand solu­tions, are the pri­ma­ry moti­va­tions for this stu­dy. In the absence of cen­tral control, we are inter­es­ted in for­ming a fleet of auto­no­mous vehicles capable of respon­ding to citys­cale tra­vel demands. Typically, this pro­blem is sol­ved cen­tral­ly ; this implies […]
  • Innovation, in eco­no­mic acti­vi­ty, in mana­ge­rial concepts and in engi­nee­ring desi­gn, results from crea­tive acti­vi­ties, entre­pre­neu­rial stra­te­gies and the busi­ness cli­mate. Innovation leads to tech­no­lo­gi­cal, orga­ni­za­tio­nal and com­mer­cial changes, due to the rela­tion­ships bet­ween enter­prises, public ins­ti­tu­tions and civil socie­ty orga­ni­za­tions. These inno­va­tion net­works create new know­ledge and contri­bute to the dis­se­mi­na­tion of new socio‐economic […]
  • Nos étu­diants sont en quête de sens, veulent être utiles…autrement dit, évo­luer dans une Université res­pon­sable ! Mais de quoi s’agit-il exac­te­ment ? Mon inter­ven­tion s’articule autour de deux points prin­ci­paux. Dans un pre­mier temps, je revien­drai sur la ques­tion du « Pourquoi » d’une Université res­pon­sable. Les éta­blis­se­ments d’enseignement supé­rieur sont inci­tés à revoir […]
  • Agent orga­ni­za­tions offer power­ful abs­trac­tions to build dis­tri­bu­ted sys­tems. Current models, howe­ver, lack of a sys­te­ma­tic way to address excep­tions as part of their desi­gn. Thus, excep­tion hand­ling is usual­ly approa­ched by ad‐​hoc solu­tions, that ham­per code modu­la­ri­za­tion and decou­pling. We present an exten­sion of the orga­ni­za­tio­nal model and infra­struc­ture adop­ted in JaCaMo, that explicitly […]
  • […]
  • Today’s fast‐​growing demands at the glo­bal level for mobile appli­ca­tions (apps) cause cus­to­mers to call for the cus­to­mi­za­tion of their apps to fit their indi­vi­dua­li­zed needs and busi­ness realms. Customization is a chal­lenge for apps‐​development com­pa­nies when they want to satis­fy their nume­rous users in a crow­ded com­pe­ti­tive mar­ket. Moreover, pur­suing cus­to­mi­za­tion involves addi­tio­nal challenges […]
  • Telemedicine is not a simple tech­no­lo­gy but a context‐​dependent inno­va­tion implying signi­fi­cant pro­cess modi­fi­ca­tions. Despite the gro­wing inter­est in tele­me­di­cine imple­men­ta­tion, it still does not form part of the heal­th­care orga­ni­za­tions’ recur­rent acti­vi­ties due to seve­ral obs­tacles. The neglect of pro­cess change is one of the renow­ned bar­riers. This work’s ori­gi­na­li­ty lies in using the […]

Nos dernières publications en texte intégral

  • Kriging is a wide­ly employed tech­nique, in par­ti­cu­lar for com­pu­ter expe­ri­ments, in machine lear­ning or in geo­sta­tis­tics. An impor­tant chal­lenge for Kriging is the com­pu­ta­tio­nal bur­den when the data set is large. This article focuses on a class of methods aiming at decrea­sing this com­pu­ta­tio­nal cost, consis­ting in aggre­ga­ting Kriging pre­dic­tors based on smal­ler data […]
  • The research Chair OQUAIDO – for « Optimisation et QUAntification d’Incertitudes pour les Données Onéreuses » in French – gathers aca­de­mic and tech­no­lo­gi­cal research part­ners to work on sta­tis­ti­cal lear­ning pro­blems invol­ving scarce and error‐​prone data. This Chair, was crea­ted in January 2016 for a per­iod of 5 years. It has tack­led pro­blems where small data is […]
  • We construct the Copula Recursive Tree (CORT) esti­ma­tor : a flexible, consistent, pie­ce­wise linear esti­ma­tor of a copu­la, leve­ra­ging the patch­work copu­la for­ma­li­za­tion and various pie­ce­wise constant den­si­ty esti­ma­tors. While the patch­work struc­ture imposes a grid, the CORT esti­ma­tor is data‐​driven and constructs the (pos­si­bly irre­gu­lar) grid recur­si­ve­ly from the data, mini­mi­zing a cho­sen dis­tance on […]
  • Agent orga­ni­za­tions offer power­ful abs­trac­tions to build dis­tri­bu­ted sys­tems. Current models, howe­ver, lack of a sys­te­ma­tic way to address excep­tions as part of their desi­gn. Thus, excep­tion hand­ling is usual­ly approa­ched by ad‐​hoc solu­tions, that ham­per code modu­la­ri­za­tion and decou­pling. We present an exten­sion of the orga­ni­za­tio­nal model and infra­struc­ture adop­ted in JaCaMo, that explicitly […]
  • Today’s fast‐​growing demands at the glo­bal level for mobile appli­ca­tions (apps) cause cus­to­mers to call for the cus­to­mi­za­tion of their apps to fit their indi­vi­dua­li­zed needs and busi­ness realms. Customization is a chal­lenge for apps‐​development com­pa­nies when they want to satis­fy their nume­rous users in a crow­ded com­pe­ti­tive mar­ket. Moreover, pur­suing cus­to­mi­za­tion involves addi­tio­nal challenges […]
  • L’utilisation crois­sante d’al­go­rithmes d’Intelligence Artificielle (IA) dans des appli­ca­tions impac­tant des humains requiert de doter ces sys­tèmes d’un com­por­te­ment pou­vant être jugé éthique selon des valeurs humaines. Bien que plu­sieurs approches existent, la ques­tion de l’a­dap­ta­tion au contexte, aux pré­fé­rences et prin­cipes éthiques des uti­li­sa­teurs, reste posée. Nous pro­po­sons de trai­ter cette ques­tion par l’Apprentissage […]
  • Cet article pré­sente le contexte et les objec­tifs du pro­jet ANR nom­mé Constrained Semantic Web of Things (CoSWoT). CoSWoT a pour objec­tif de pro­po­ser une archi­tec­ture logi­cielle dis­tri­buée com­pa­tible WoT et embar­quée sur des dis­po­si­tifs contraints. Cette archi­tec­ture a deux carac­té­ris­tiques prin­ci­pales : (1) elle uti­li­se­ra des modèles de connais­sances à base de graphes pour […]
  • […]
  • Dans le cadre du trans­port à la demande (TAD), l’ap­pa­rie­ment cen­tra­li­sé tra­di­tion­nel per­met d’ob­te­nir des solu­tions opti­males, mais reste NP‐​difficile et donc inadap­té aux pro­blèmes en ligne et dyna­miques. Les approches décen­tra­li­sées per­mettent d’ob­te­nir des solu­tions réa­li­sables en temps réel, mais sans garan­tie de qua­li­té. Ici, nous consi­dé­rons une ver­sion dyna­mique du TAD, appe­lée Dial‐A‐Ride‐Problem […]
  • Allocation pro­blems are major issues in mana­ging On‐​Demand Transport (ODT) sys­tems. They have been stu­died for decades, and a varie­ty of solu­tions were pro­po­sed. The approaches to these pro­blems can be clas­si­fied as cen­tra­li­zed and decen­tra­li­zed. Each has its pros and cons in prac­tice. In this work, we aim to pro­vide a gene­ric model for […]