STIC, MATH & CLIMAT AmSud

CLIMAT AmSud project with LNCC

 The Green AI project’s main goal is to conceive a systemic and multi-component approach to the problem of the Artificial Intelligence’s ecological impact. Thus, it focuses on cloud and mobile computing, transfer learning, model reuse, active learning and evolutionary computing, among others. “Green AI: towards an ecologically viable machine learning” was born in 2019 from an initiative of Inria Chile in partnership with PUC-Rio. Researchers from LNCC under the coordination of Mariza Ferro joined the project, as well as researchers from other countries CMM (Centro de Modelamento Matematico de Chile), Universidad de la Republica (Uruguay) and Universidad de Asuncion (Paraguay). This project was selected in the frame of the 1st Call for projects CLIMAT AmSud and receives fundings from Inria, ANID and Conacyt.

STIC AmSud projects with LNCC

The MOTIf project, funded by the Stic-AmSudprogram, is a two-years long international project (2018-2019) between French and Latin American partners from Argentina and Brazil. The members of this project are Inria (DANTE and INFINE teams) and ENS-Lyon in France, the UBA (Universidad de Buenos Aires) and Grandata in Argentina, LNCC,  UFMG (Universidade Federal de Minas Gerais) and PUCMG (Pontifícia Universidade Católica de Minas Gerais) in Brazil. The project had the aim of, starting from mining and analyzing mobile data, developing prediction models of human behavior, identifying correlations between demographic and socioeconomic attributes, and investigating the evolution of the social network that emerges from communication between people.

UCOOL (Understanding and predicting human demanded COntent and mObiLity) is the STIC-AmSud project which gathered LNCC in Brazil, UBA (Universidad de Buenos Aires, Faculty of Engineering, Argentina), USM (Universidad Tecnica Federico Santa Maria, Chile) Telecom Sud Paris and Inria. It lasted from 2014 to 2015. The project aimed at defining solutions for the identification and modelling of correlations between the user mobility – describing changes in the user positioning and the current environment he/she is in – and the traffic demand he/she generates. It also had as an objective to develop techniques to predict the future user mobility and demanded content at both microscopic (i.e., individual user) and macroscopic (i.e., user communities) levels. The insights obtained from these studies should drive the proposal of new content- and context-aware networking protocols and/or applications that will explicitly leverage the mobility and interest of users to improve the perceived quality of their services and consequently, improve user satisfaction.

Math AmSud projects with LNCC

The EOLIS Math AmSud project (Efficient Off-LIne numerical Strategies for multi-query problems) will start on January 2021 for a two years duration. It involves Inria, represented by Théophile Chaumont from Atlantis Project-Team, LNCC with the PI Frederic Valentin and UdeC (Universidad de Concepcion, Chile) featuring the researcher Manuel Solano. This work targets multi-query problems that involve the solution of a parametrized boundary value problem (BVP) for a set of parameters. Notable examples include parameter identification, optimal design and uncertainty quantification, each having concrete applications.

PHOTOM project (Photovoltaic Solar Devices in Multiscale Computational Simulations), 2018-2019 was conducted  within the framework of the MATH AmSud program during two years collaboration between Inria, LNCC (Brazil), UDEC (Universidad de Concepcion, Chile) and PUCV (Pontificia Universidad Catolica, Chile). Its purpose was to contribute at “developing, analyzing and implementing innovative multiscale finite element numerical methods for was propagation models in grating media motivated by its use in the simulation of photovoltaic solar cells”.

To know more about STIC and MATH AmSud program, please follow this link.