Louai Saker

Ensuring fairness in machine learning while maintaining predictive performance remains a fundamental challenge in data science. Most fairness-aware learning approaches rely on fixed penalty scalarization or static multi-objective formulations, which often lead to unstable trade-offs and sensitivity to manually tuned hyperparameters. In this paper, we propose SAFEA (Self-Adaptive Fairness Entropy Algorithm), a novel evolutionary optimization framework that dynamically regulates the fairness–accuracy trade-off using inequality-aware feedback mechanisms. SAFEA introduces two complementary measures: the Fairness Entropy Index (FEI), which captures the dispersion of group-level fairness violations, and the Gini Fairness Index, which quantifies disparity in prediction errors across protected groups. These measures guide an adaptive penalty update rule that autonomously adjusts the fairness coefficient during the evolutionary search process, eliminating the need for manual tuning. Theoretical analysis establishes boundedness and stability of the adaptive penalty under mild assumptions and discusses convergence properties under Lipschitz-continuous objectives. Experimental evaluation on benchmark datasets (Adult Income, COMPAS, and German Credit) demonstrates that SAFEA improves hypervolume by up to 12.4% compared to NSGA-II fairness formulations, reduces demographic parity difference by 18–25% relative to static penalty evolutionary methods, and achieves up to 3.1% higher F1-score than adversarial debiasing approaches while maintaining competitive accuracy. These results indicate that entropy-guided adaptive regulation leads to smoother fairness convergence and better Pareto front coverage. The proposed framework bridges inequality theory and evolutionary multi-objective optimization, providing a scalable and effective solution for fairness-aware learning in high-stakes applications.

Mécanismes de réduction de la consommation d'énergie dans les réseaux mobiles: La consommation en énergie des réseaux de télécommunications a suscité un intérêt grandissant au cours des dernières années et les opérateurs mobiles cherchent des solutions innovantes pour optimiser l’efficacité énergétique. Dans cette thèse, nous nous focaliserons sur les schémas d’optimisation de la consommation d’énergie des réseaux d’accès mobiles. Nous commençons par étudier la stratégie de partage de charge qui minimise la consommation d’énergie dans la cellule tout en limitant l’impact sur la Qualité de Service. Le gain obtenu par ces algorithmes de gestion des ressources radio reste cependant limité; ceci est dû au fait que la consommation d’énergie ne dépend pas que de la charge, mais comporte une importante partie constante. Pour obtenir un gain plus important, nous proposons la mise en veille de certaines ressources du réseaux aux heures creuses où le trafic est faible, ce qui mène à des gains substantiels. Nous proposons ensuite un contrôleur de gestion de la mise en veille qui choisit l’action optimale en fonction de l’état du réseau. En effet, l’activation d’une nouvelle ressource n’est pas instantanée et un effet ping-pong peut apparaitre suite aux commandes simultanées d’activation/désactivation des ressources. Nous adaptons nos contrôleurs afin de prendre en compte ces imperfections, et nous montrons comment déduire la politique optimale. Nous étudions ensuite le cas de déploiement de petites cellules dans des réseaux hétérogènes et montrons que leur efficacité énergétique est améliorée par rapport au réseau purement macro-cellulaire, pourvu que la consommation d’énergie de petites cellules reste faible. Nous proposons ensuite un contrôleur optimal qui active/désactive les petites cellules en se basant sur des informations de trafic et de localisation des usagers. Nous considérons différents cas de figure avec une information complète, partielle ou retardée et montrons que ces schémas de mise en veille permettent d’atteindre d’importants gains de consommation d’énergie

Efficient and timely transportation of blood samples across medical centers is critically important, particularly in countries where traffic congestion significantly delays road-based delivery. The operational range of the Unmanned Aerial Vehicles remains constrained by limited battery capacity and variable energy consumption across different terrains. This paper proposes an integrated approach for optimizing drone-based medical transportation by combining a 3D enhanced A* pathfinding algorithm with a K-Means clustering method for strategic deployment of drone charging stations. The 3D A* algorithm incorporates elevation and environmental obstacles, enabling accurate estimation of flight costs. Charging 1 station placement considers population density, topography, power infrastructure , and drone power consumption models. A case study involving Lebanese hospitals demonstrates the effectiveness of the proposed system in minimizing flight distance, reducing energy consumption, and ensuring mission feasibility under real-world constraints.

This paper proposes a novel Hybrid Metaheuristic Framework (HMF) that integrates multiple optimization strategies within a self-adaptive mechanism for solving large-scale combinatorial problems. Unlike traditional metaheuristics, which rely on fixed algorithmic parameters and predefined strategies, HMF dynamically selects and adjusts heuristic operators using a reinforcement learning-based adaptive mechanism. Specifically, the framework employs Deep Q-Learning to guide real-time parameter tuning and metaheuristic selection based on performance feedback from the optimization process. The study applies this framework to three benchmark problems: the Vehicle Routing Problem (VRP), the Facility Location Problem (FLP), and the Job Scheduling Problem (JSP), demonstrating consistent gains in solution quality and convergence speed. Experimental results reveal that HMF achieves noticeable improvements over conventional methods in both runtime efficiency and optimization effectiveness, without requiring manual parameter tuning. The findings establish the proposed framework as a promising tool for real-world large scale combinatorial optimization challenges.

Disaster logistics presents a highly complex decision-making challenge under conditions of uncertainty, where the timely and efficient allocation of scarce resources is essential to minimize human suffering. In this context, we propose a novel Quantum-Inspired Hyperheuristic Framework (QHHF) designed to solve Dynamic Multi-Objective Combinatorial Optimization Problems (DMOCOPs) arising in disaster relief operations. The proposed framework integrates Quantum-Inspired Evolutionary Algorithms (QIEAs), which facilitate diverse and explorative solution generation, with a Reinforcement Learning (RL)-based hyperheuristic capable of dynamically selecting the most suitable low-level heuristic in response to evolving disaster conditions. A dynamic multi-objective mathematical model is formulated to simultaneously minimize total travel cost and risk exposure, while maximizing priority-weighted demand satisfaction. The model captures real-world complexity through time-dependent variables, stochastic demand variations, and fluctuating transportation risks. Extensive simulations using real-world disaster scenarios demonstrate the effectiveness of the proposed approach in generating high-quality solutions within stringent response time constraints. Comparative evaluations reveal that QHHF consistently outperforms traditional heuristics and metaheuristics in terms of adaptability, scalability, and solution quality across multiple objective trade-offs. Notably, our method achieves a 9.6% reduction in total travel cost, a 6.5% decrease in cumulative risk exposure, and a 4.7% increase in priority-weighted demand satisfaction when benchmarked against existing techniques. This work contributes both to the advancement of hyperheuristic theory and to the development of practical, AI-enabled decision-support tools for emergency logistics management.

Properly managing waste generated by buildings and public works is a significant challenge in Tunisia, particularly in the city of Bizerte. The inadequate disposal of such waste can cause substantial harm to human life, property, and the environment. This paper proposes an multi-criteria decision making (MCDM) that utilizes the analytic hierarchy process (AHP) decision support tool to identify suitable landfill sites for construction and demolition waste (CDW) in Bizerte. The AHP method is widely used in MCDM applications. The approach involves classifying different scenarios based on various exclusion and appreciation criteria to determine the optimal locations for future landfills. Furthermore, the paper develops a conceptual approach for identifying better sites for the disposal of CDW, resulting in a comprehensive database capable of storing, accessing, and extracting information at both conceptual and operational levels. The proposed model considers spatial, technical, and environmental criteria in the selection of a suitable landfill site. The proposed methodology offers an effective and practical solution for properly managing CDW waste in Bizerte, Tunisia, and can be applied to other regions facing similar challenges.

The agricultural sector is undergoing a transformative paradigm shift with the integration of advanced technologies, particularly artificial intelligence (AI), to enhance data analysis techniques and streamline decision-making processes. This paper delves into the integration of advanced technologies in agriculture, focusing specifically on optimizing data analysis through artificial intelligence (AI) to strengthen decision-making processes in farming. We present a novel AI-powered model that leverages historical agricultural datasets, utilizing a comprehensive array of established machine learning algorithms to enhance the prediction and classification of agricultural data. This work provides tailored algorithm recommendations, bypassing the need to deploy and fine-tune numerous algorithms. We approximate the accuracy of suitable algorithms, highlighting those with the highest precision, thus saving time by leveraging pre-trained AI models on historical agricultural data. Our method involves three phases: collecting diverse agricultural datasets, applying multiple classifiers, and documenting their accuracy. This information is stored in a CSV file, which is then used by AI classifiers to predict the accuracy of new, unseen datasets. By evaluating feature information and various data segmentations, we recommend the configuration that achieves the highest accuracy. This approach eliminates the need for exhaustive algorithm reruns, relying on pre-trained models to estimate outcomes based on dataset characteristics. Our experimentation spans various configurations, including different training–testing splits and feature sets across multiple dataset sizes, meticulously evaluated through key performance metrics such as accuracy, precision, recall, and F-measure. The experimental results underscore the efficiency of our model, with significant improvements in predictive accuracy and resource utilization, demonstrated through comparative performance analysis against traditional methods. This paper highlights the superiority of the proposed model in its ability to systematically determine the most effective algorithm for specific agricultural data types, thus optimizing computational resources and improving the scalability of smart farming solutions. The results reveal that the proposed system can accurately predict a near-optimal machine learning algorithm and data structure for crop data with an accuracy of 89.38%, 87.61%, and 84.27% for decision tree, random forest, and random tree algorithms, respectively.