AI Research: EMSI Hosts International Experts in Casablanca for ICOMP 2026

Casablanca hosted the third edition of the International Conference on Computational Optimization (ICOMP 2026) from November 3 to 6, 2026. Organized for the first time on the African continent, this international scientific meeting brought together researchers and specialists in artificial intelligence to discuss advancements in computational optimization. For the Moroccan School of Engineering Sciences (EMSI), this event represents a new opportunity to strengthen its positioning within international research networks.

Casablanca, a New Milestone in the International Scientific Agenda

Over four days, the EMSI Casa Anfa campus welcomed academics and experts to exchange ideas on the mathematical and algorithmic foundations of artificial intelligence. The conference was organized by EMSI through its Multidisciplinary Research and Innovation Laboratory (LPRI), in collaboration with Innopolis University and the Mohamed Bin Zayed University of Artificial Intelligence (MBZUAI).

After a first edition in Innopolis, Russia, in 2024, and a second in Abu Dhabi, UAE, in 2025, ICOMP chose Casablanca to continue its international development. This first organization in Africa reflects the conference's openness to new scientific ecosystems and provides Morocco with additional visibility in discussions dedicated to advanced technologies.

According to Dr. Mohamed Tabaa, director of LPRI, hosting this event is an important milestone for EMSI and its laboratory. He particularly emphasizes the importance of strengthening international scientific collaborations and enabling Moroccan students and young researchers to meet world-renowned specialists.

International Researchers in Attendance

The ICOMP 2026 program featured several prominent figures in the fields of artificial intelligence, machine learning, and mathematical optimization. Among the announced speakers was Michael Jordan, a professor at the University of California, Berkeley, and a leading figure in statistical learning and artificial intelligence.

The conference also included Konstantin Mishchenko, presented as a researcher in artificial intelligence at Mistral and a specialist in optimization methods applied to machine learning. Other researchers contributed to the program, including Antonio Orvieto from the ELLIS Institute Tübingen and the Max Planck Institute for Intelligent Systems, Anatoli Juditsky from the University of Grenoble Alpes, Anastasia Koloskova from the University of Zurich, Ajanthan Thalaiyasingam from the Australian National University, and Peter Richtárik from the King Abdullah University of Science and Technology (KAUST).

The diversity of the represented institutions illustrates the purpose of this meeting: to bring fundamental research in applied mathematics closer to scientific and technological issues related to the development of artificial intelligence.

Computational Optimization Facing AI Challenges

Beyond the visible performance of artificial intelligence models, their development relies on mathematical and algorithmic methods that determine their learning capacity and computational resource needs.

In a context where models are becoming increasingly complex, computational optimization becomes a central issue. Improving calculation efficiency, reducing training costs, and accelerating algorithm execution are among the challenges addressed during the conference.

The presented work focused on efficient training and inference of models, new neural architectures, model compression, distributed algorithms, and large-scale optimization methods. These research areas aim to enhance the performance of AI systems while minimizing the resources required for their operation.

Potential applications span various fields, from industry to digital services, embedded systems, and scientific applications.

A Focus on Students and Young Researchers

ICOMP 2026 also integrated an educational dimension through ASCOMP 2026, a scientific school associated with the conference dedicated to computational optimization and artificial intelligence.

Targeted primarily at master's students, doctoral candidates, and young researchers, this initiative offered specialized teachings, exchanges with international researchers, and sessions for presenting scientific work. It thus provided participants with a framework to deepen their knowledge and compare their approaches with work conducted at other institutions.

For Moroccan students, holding these activities in Casablanca represents an opportunity to directly access international expertise and better understand the research directions in artificial intelligence.

EMSI Focuses on Scientific Partnerships

Through the organization of this third edition, EMSI aims to solidify the role of research in its academic development. The involvement of its Multidisciplinary Research and Innovation Laboratory, led by Dr. Mohamed Tabaa, aligns with this direction, with a particular interest in artificial intelligence, optimization, and applied mathematics.

Founded in 1986 and recognized by the state, EMSI trains engineers in various fields, including digital technology, industry, civil engineering, finance, and intelligent systems. A member of Honoris United Universities, the institution has 21 campuses spread across six cities in Morocco, as well as a campus in Nouakchott, Mauritania. It hosts over 18,000 students and claims a network of more than 26,000 active alumni in Morocco and internationally.

By hosting ICOMP 2026, the school seeks to strengthen the links between engineering education, fundamental research, and international collaborations in a sector where mastery of mathematical and computational tools is increasingly important.