Research teams
Computer Science and Artificial Intelligence
IIA
- Head
- To be appointed
- Staff
- membership being entered
The IIA team answers the question: how is it learned, executed and deployed?
Overview
Learning and computing systems applied to scientific and industrial data.
Research topics
I1 — Machine learning and deep learning
Supervised and unsupervised learning, deep neural networks, reinforcement learning, learning from limited or imbalanced data, model evaluation and robustness.
I2 — Data science and heterogeneous data
Environmental, biomedical and industrial data; remote sensing, time series, source fusion, data quality and completeness, preparation and documentation of datasets.
I3 — Computer vision and pattern recognition
Image processing and analysis, segmentation, handwritten Arabic character recognition, imaging applied to environment and health.
I4 — Scientific computing, software engineering and systems
High-performance computing and parallelisation, reproducibility and version control, databases and scientific information systems, IoT and embedded systems, digital twins, monitoring and data acquisition.
The IIA team provides the unit's shared technical foundation: computing environments, data and code repositories.
Contribution to the thematic axes
| Axis | Team's work |
|---|---|
| ERN | Remote sensing and environmental monitoring, environmental databases, processing pipelines, high-performance computing applied to simulation |
| SPE | Biomedical data analysis, health information systems, decision-support tools |
| GIOSI | Industry 4.0, predictive maintenance, anomaly detection, digital twins, process monitoring |
Skills and tools
Python, PyTorch, TensorFlow, scikit-learn, SQL, Git, parallel computing, Arduino and Raspberry Pi.
Work carried out with the MSM team
Optimisation of learning architectures; uncertainty quantification of predictions; coupling of physical models and learned models.
Members
The team's membership will be shown as soon as members have been assigned to it in the researcher area.
Contact
For any question about the team, write to contact@umr-ames.mr