Author(s):
Quintas, José Miguel Rodrigues
Date: 2025
Persistent ID: http://hdl.handle.net/10362/190414
Origin: Repositório Institucional da UNL
Subject(s): maritime tourism; resource allocation; genetic algorithms; NSGA-II; multi-objective optimization; single-objective optimization; evolutionary algorithms; metaheuristics; combinatorial optimization; constraint satisfaction; multi-skilled resource-constrained project scheduling; Design Science Research Methodology; comparative analysis; statistical validation; operations research; scheduling optimization; Pareto optimization; weighted sum aggregation; domain-specific genetic operators; constraint handling; tour planning; guide assignment; vessel allocation; skipper scheduling; tourism operations; service optimization; operational efficiency; revenue optimization; real-world optimization; empirical evaluation; performance comparison; computational efficiency; solution quality; feasibility analysis; crowding distance; diversity maintenance; population-based optimization; stochastic optimization; optimization algorithms; resource utilization; capacity constraints; language compatibility; multilingual services; operational research; applied optimization; algorithmic comparison; strategic planning; emergency response; decision support systems; SDG 8 - Decent work and economic growth; SDG 9 - Industry, innovation and infrastructure; SDG 12 - Responsible production and consumption; SDG 14 - Life below water; Domínio/Área Científica::Ciências Naturais::Ciências da Computação e da Informação
Description
Dissertation presented as the partial requirement for obtaining a Master's degree in Data Science and Advanced Analytics, specialization in Data Science
Maritime tourism resource allocation requires simultaneous assignment of multilingual guides, vessels, and qualified skippers to customer reservations while satisfying multiple operational constraints and competing objectives. Despite its operational importance, existing literature lacks rigorous comparative studies evaluating single-objective versus multiobjective optimization approaches using real-world data. This thesis presents the first systematic comparative analysis of genetic algorithm approaches for maritime tourism resource allocation, comparing a Custom GA (single-objective with weighted sum aggregation) against a Custom NSGA-II (multi-objective with Pareto dominance). Both algorithms employ identical problem representations, constraint handling mechanisms, and domain-specific genetic operators while differing in optimization philosophy. The evaluation utilizes authentic operational data from an active maritime tourism company, comprising 83 reservations, 21 multilingual guides, 15 vessels, and 11 qualified skippers. The empirical evaluation employs 120 independent trials (60 per algorithm) with rigorous statistical validation including appropriate test selection, effect size calculation, and multiple comparison corrections. The Custom NSGA-II extends traditional multi-objective optimization to five dimensions, representing a significant theoretical advancement in high-dimensional optimization. Results demonstrate that Custom NSGA-II achieves superior overall performance with 3.1% higher fitness scores (p < 0.001) and exceptional constraint satisfaction (99.97% vs 53.07% feasible solutions, p < 0.001), while Custom GA provides 3.7× faster execution (p < 0.001). Most significantly, both algorithms achieved perfect resource allocation (83/83 reservations) compared to human planners' maximum of 76/83 reservations, representing a concrete 9.2% improvement that translates directly to revenue enhancement. The research contributes novel domain-specific genetic operators, the first five-objective NSGA-II implementation for resource allocation, and comprehensive statistical validation methodology. Practical contributions include evidence-based algorithmic selection guidance, with NSGA-II recommended for strategic planning (90% of cases) and Custom GA for emergency response (10% of cases). The findings demonstrate that sophisticated optimization techniques can transform complex operational challenges into competitive advantages through enhanced efficiency and improved service quality.
