Apply for a job
|H||Healthcare analyticsHPC resources||HPC accounts||HPC centre|
|I||ISCD members||ISCD missions|
|L||Labex description||Labex forms|
|S||Scientific Advisory BoardStaff
|Scientific initiativesSteering Committee||SimseoSummer school|
|T||Technical support||Training catalogue||Transportation|
|U||Usage policies HPC||Usage policy meeting room||User committee|
The research at the ISCD is dedicated to developing and promoting mathematical and computational methods across sciences, humanities and medicine. This involves defining and analyzing new mathematical models, designing new computer algorithms and tailored software for efficient high-performance computing. The institute has acquired a unique expertise in scientific computing and data analysis by gathering scientists from various disciplines. These scientists work together within the ISCD on interdisciplinary projects, but are also part of their home faculties. In this spirit, a right balance has been achieved between disciplinary excellence and interdisciplinary cooperation.
The key assets of the ISCD include: large scale computations, big data analysis and scientific visualization. In a few years, the ISCD has become one of the leading interdisciplinary research centers of Sorbonne Université with a dozen of active members and several research groups from all three faculties: sciences, humanities and medicine, spanning a large range of disciplines, i.e. mathematics, computer sciences, biology, chemistry, physics, earth sciences, medicine, anthropology, archaeology, linguistics, philosophy, ethics, etc.
To tackle challenging problems more effectively, some interdisciplinary collaborations have been set-up and organized around project-teams. A project-team is structured around a multidisciplinary team of scientists sharing a research program and objectives under the guidance of a team leader. Each research project has a large scientific independence, benefit from resources allocated by the institute and by other initiatives. Its evaluation is carried out every four years by the members of the Scientific Advisory Committee.
Examples of domains in which challenging problems have been addressed by multidisciplinary research teams include:
Computational Methods for (meta)Genomics
The team works on various problems connected with the study of the functioning and the evolution of biological systems, based on “omics” data. Molecular modeling, machine learning, statistics and combinatorics are heavily involved in this challenging field.
Mathematics for Computational Chemistry
Significant breakthroughs have been obtained in the description of theoretical models, the analysis of numerical schemes and the efficient computer implementation along three main axis: electronic structure, multiscale quantum modeling and chemical interpretation.
In addition, junior groups, partly located on the premises of the ISCD, are investigating new areas of applications in personalized health care, digital humanities, digital archaeology, computational forensics sciences or scientific visualization.
Throughout these programs, the ISCD has acquired a valuable expertise in scientific computing, mathematical modelling and data analysis. Fresh programmes based on updated and state-of-the-art models that are both efficient for new HPC architectures and effective for advancing high-level research have been developed and a large group of people has the necessary competency to use and to upgrade them. This is also initiating a new cycle in the development of industrial applications. Core research fields include:
Multilevel, domain decomposition and parallel methods
The emergence of parallel computers and their potential for solving complex multi-physics problems had led to a large amount of research and development in domain decomposition methods, multilevel and multigrid algorithms.
Optimization and Control Techniques
In the last decade, extensive research work has been dedicated to exploring efficient mathematical and numerical methods for solving complex optimization problems, like optimal design, shape optimization, optimal control.
Data Analysis and Statistical Methods
Statistics and data analysis is a very active field of research and development, which is driven by the pressing demand in analysis and modelling of complex systems and large datasets, in many disciplines.
Machine Learning Algorithms
The field of machine learning and artificial intelligence has gained more popularity and interest in the past couple of years. Our computer scientists explore the study and the design of efficient algorithms that can learn from and make prediction on data.
HPC Methods for Applications
Experts and engineers are working together to develop and share methodologies, numerical methods and their implementation used by the state-of-the-art codes in the HPC environment. Computational tools (visualisation and parallel computing) are made available to a large scientific community.
Computing at extreme scales or encompassing many scales on a wide spectrum is not confined to computational chemistry or biology. The same applies to sequential algorithms and to massive data analysis. Therefore, with more financial support expected, the ISCD approach could easily transfer and adapt to build more packages in other disciplines, in particular:
Seminars and symposia are organized to strenghtening these research areas and to identify leading experts and promising young investigators in these application fields. Such interactions will in turn lead to submitting proposals for funding by a small group of researchers. The ISCD will contribute to establish further bridges between research domains by promoting mathematical modelling and scientific computing to disciplines where they are not fully exploited.