Accueil
Artificial Intelligence and Complex Systems
AICS in the Complex System Digital Campus (CS-DC)
Learn more on the CS-DC on Cosypedia.
The UN Sustainability Goals
AICS actively contribute to research in the domain of UN Sustainability Goals:
- 3: Health
- 4: Education
- 9: Industry
Link with CS-DC e-laboratory
AICS members are responsible for two CS-DC e-laboratories:
- E-Laboratory on complex computational ecosystems (ECCE)
- E-Laboratory on the Factory of the Future (4P-Factory)
Contribution to CS-DC Flagships
AICS members actively contribute to following CS-DC Flagships:
- Reconstruction And Prediction of multi-Scale Overall Dynamics (Rapsody)
- Personalized Health phYSIcally, sOcially and Mentally for Each in their networkS (Physiome)
- Personalized Open Education for the Masses (POEM)
Organisation of AICS
The AICS is part of Campus Numérique des Systèmes Complexes de l'Université de Strasbourg.
The AICS is organised around 3 poles:
- Foundations
- Quantum algorithms
- Explainable Artificial Intelligence
- Immune Ecosystems
- Services
- Infrastructure Services
- AICS-Data
- Secure 4P Ecosystems
- Applications
- Health
- Education
- Computation
- Industry
Foundation of Artificial Intelligence for Complex Systems
Quantum algorithms
Approaches:
- Quantum-inspired algorithms
- Algorithms for quantum computers
Members: Pierre Collet, Pierre Parrend, Anna Leontova
Explainable Artificial Intelligence
Approaches:
- Design and optimisation of 4P algorithms
- Explainable AI
Members: Anne Jeannin-Girardon, Pierre Collet, Pierre Parrend, Romain Ohrand, Naser Ghannad
Immune Ecosystems
Approaches:
- Design of data handling chains for anomaly detection
- IA solutions for cybersecurity
Members: Pierre Parrend, Rabih Amhaz, Aline Deruyver, Julio Navarro, Fabio Guigou
Support
Infrastructure Services
Services:
- Evolutionary computing
- Data and log storage
- Applicative maintenance
- Digital courses hosting
Members: Arnaud Kress, Pierre Collet, Raymond Ripp
AICS-Data
Services:
- Data integration ('Data Steward' role), in particular in data preparation (annotation)
- Characterisation of data quality, to support automatised analysis
Members: Arnaud Kress, Laetitia Poidevin, Luc Moulinier, Julie Thompson, Pierre Parrend
Secure 4P Ecosystems
Services:
- Data protection: RGPD compliance; securisation of data flows
- Audit and design of secure applications
- Cohort and databases management
Members: Pierre Parrend, Pierre Collet, Rabih Amhaz
Domains of AI-CS
Health
Education
Industry
Resources
Infrastructure
- Parsec for parallel GPGPU evolutionary computation
- Deployed in: Strasbourg, Rouen (France), Nouakchott (Mauritania)
- 4.0 Security Rack
- Deployed in: ECAM Strasbourg-Europe, Schiltigheim, France
Software
- EASEA (Easy Specification of Evolutionary Algorithms) & EASEA-Cloud : Massively parallel computing platform(300 TFlops) dedicated to artificial evolution, deep learning and multi-layer neural networks on GPGPU machines.
- Web : EASEA (Easy Specification of Evolutionary Algorithms)
- Contact : Pierre Collet, E-mail : pierre.collet[AT]unistra.fr
- HOPE (Health Optimisation through 4P Ecosystems) : Patient cohort management platform which supports integration, analysis and exploitation of massive medical cohorts (more than 500.000 patients) or rare diseases (some dozens of patients), compliant with European GDPR regulation.
- Web: HOPE (Health Optimisation through 4P Ecosystems)
- Contact: Pierre Parrend, E-mail: pierre.parrend[AT]unistra.fr
- RADAR : A distributed environment for secure, robust and anonymous data storage, which supports request and analysis, in particular for medical data.
- Web: RADAR (Robust Anonymous Data Records)
- Contact: Rabih Amhaz, rabih.amhaz[AT]unistra.fr
- POEM (Personalized Open Education for the Masses) : A social and intelligent plateform for a massive and personnalised educationnal ecosystem.
- Web : POEM (Personalized Open Education for the Masses)
- Contact : Pierre Collet, E-mail : pierre.collet[AT]unistra.fr
- ClowdFlows : Unique web platform which supports execution and comparison, of relational data mining techniques.
- Site web : ClowdFlows
- Contact : Nicolas Lachiche, E-mail : nicolas.lachiche[AT]unistra.fr
Projects
80|Prime
Virtfac
- Web: VirtFac Project
Halfback
- Web: Halfback Project