Keynote Speakers

Abdelhamid Benaini
Professor of Computer Sciences, Normandie University Le Havre
Machine learning algorithms for combinatorial optimization problems

Renaud Di francesco

PhD, Fellow of the Institution of Engineering and Technology, member of the British Computer Society
Disrupting data-vector-based operational models: -Matrices of sets for heterogeneous data aggregated in observation sets with candidate use cases
A framework for Matrices of Sets, will be presented, at the convergence of the 1850 matrix theory and the 1870 theory of sets, both well-known, with two candidates for the generalisation of the large summation operator of classical matrix products and its term-to-term product.

Matrices of set borrow -the efficiency and possible factorisation/reduction of matrix computations -the smooth capture by observation sets, of heterogeneity, irregularity and incompleteness of available observations, which vector models fail to describe The couple (large operator Sigma, term-to-term product .) is taken to be either (union of sets, Cartesian product of sets) or (disjunct union of sets, intersection of sets) Merits and properties of each will be presented. In particular some theorems on eigenvalue and eigenvector properties, and other product simplification will be presented for the case of (U,x). Operational use cases generalising classical domains of application of classical matrices and vector data will be given: -graph theory and graph composition, with generalised adjacency matrices -collaborative filtering and recommendation engines (also applied in drug target repurposing) -use cases with matrices of intervals, application to monthly temperatures intervals of cities -generalised Jaccard similarity

Ihssan EL Ouadi
Professor researcher at Hight National School of Mine of Rabat
The Mathematical Programming for Agriculture, Water, and Climate Change Policies: Positive Mathematical Programming (PMP) application
Climate change has undeniable effects on agriculture, natural resources, businesses, and markets, necessitating informed decision-making regarding adaptation and mitigation strategies. Mathematical programming offers a means to address the economic dimensions of these challenges, enabling analysis even as climate change introduces unprecedented conditions. This conference aims to provide an overview of modeling techniques, conceptual considerations, and key assumptions involved in using mathematical programming to inform climate change policies. Additionally, I explore the application of positive mathematical programming (PMP) in assessing climate change impacts, adaptation, and mitigation strategies within agricultural and natural resource contexts, highlighting its policy-oriented nature. Lastly, we briefly discuss research needs in this area.

Samuel Deleplanque

Associate Professor, JUNIA-ISEN/IEMN Lille
Combinatorial Optimization and Operations Research with quantum analog machines

This topic investigates the capabilities of analog quantum computing devices from various institutions, including D-Wave in Canada and Pasqal in France. These devices perform well in solving binary, quadratic, and unconstrained optimization problems, collectively known as Quadratic Unconstrained Binary Optimization, thereby sparking interest in their application within the scientific community. Initially, this work elucidates the operational mechanisms of these machines from a computer science perspective, followed by a discussion on universal gate quantum computers, exemplified by those developed by IBM. The research work proceeds to examine a range of optimization and operations research problems that are effectively tackled using analog quantum technologies.

Among the challenges explored are the Traveling Salesman Problem (TSP), Capacitated Vehicle Routing Problem (CVRP), Job Shop Scheduling Problem (JSSP), Resource-Constrained Project Scheduling Problem (RCPSP), Max Cut, and 3-Satisfiability Problem (3-Sat). Notably, the research demonstrates that for the 3-Sat problem, employing a polynomial-time reduction to the Maximum Independent Set (MIS) simplifies the problem-solving process by generating a new, though sparser, graph that contains an increased number of variables. The study emphasizes the importance of considering specific attributes of these quantum devices, such as their incomplete qubit connectivity (D-Wave), which plays a critical role in problem-solving efficacy and implementation strategies.

Online user: 4

Privacy