Project Title: Optimizing Combinatorial Problems - Comparative Analysis of Memetic Algorithm, 
               VNS, and PSO with a Hybrid Swarm-Informed Memetic Algorithm  
Student: Yingxiao Huo 
Course: BSc in Computer Science with Artificial Intelligence  

Abstract: 
Heuristic algorithms have been widely studied and used in modern times. This study focuses on the 
further development of existing search and optimization algorithms for solving combinatorial 
optimization problems. The study conducts a comprehensive analysis of three widely-used optimization 
algorithms: Memetic Algorithm (MA), Variable Neighborhood Search (VNS), and Particle Swarm 
Optimization (PSO), examining their performance and characteristics in solving Knapscak problem 
and Generalised Assignment Problem with restricted workload distribution. The analysis encompasses 
the instability of the Memetic Algorithm , the escalating computational costs encountered by VNS 
when addressing large problems, and the factors contributing to PSO being unsuitable for this 
particular version of the GAP. Based on the findings, a novel hybrid algorithm is proposed, combining 
the strengths of the Memetic Algorithm and Particle Swarm Optimization to achieve improved solution 
quality. Upon evaluation, the novel algorithm effectively integrates the strengths of both the 
Memetic Algorithm and PSO, resulting in substantial enhancements to the GAP resolution, particularly 
in the context of addressing large-scale problems. Additionally, this study includes a thorough 
examination of the improved algorithm’s limitations and delineates the direction of further improvement.