Post by EL MANSOURI AYA

Every year, researchers introduce new metaheuristic optimization algorithms, each demonstrating strong performance on a particular class of problems. An interesting challenge is to understand whether these algorithms can be transferred to different optimization domains while preserving their effectiveness. In this project, I reproduced the Lichtenberg Algorithm, a recent physics-inspired metaheuristic originally developed for continuous optimization, and adapted it to the Traveling Salesman Problem (TSP). The proposed framework combines: - Lichtenberg Algorithm - Random Keys Encoding - 2-opt Local Search - Iterated Local Search Beyond implementing the algorithm, the project gave me the opportunity to explore scientific reproducibility, analyze the limitations of transferring optimization methods between different search spaces, and evaluate the contribution of each component through experimental comparisons. The complete implementation and technical report are available on GitHub: 🔗 https://lnkd.in/eDJygUyj I'm always happy to receive feedback or discuss possible improvements. #Optimization #Algorithms #ArtificialIntelligence #Research #ComputerScience