Thesis (BS Computer Science) -- University of the Philippines Mindanao, 2000

Many real world optimization problems are represented by Integer Programming (IP) models to derive optimal solutions. Specific problem constraints must also be satisfied by these solutions. Furthermore, the range of optimal solutions concerning only discrete, integer values is the constraint satisfied by IP models. IP problems are commonly solved using the Branch-and-Bound technique, which ultimately returns the best solutions. However, a need for better and faster search for solutions was imperative. Pure Particle Swarm Optimization (PSO) showed to be a reliable alternative to the search for optimal solutions in IP problems as presented by previous studies. With the advent of hybridization techniques to create efficient algorithms, pure PSO was paired to several other heuristics and these hybrids were used to solve many optimization problems. Hence, a fast, intelligent heuristic. Tabu Search (TS), was introduced to the pure PSO to solve IP problems. With embedding hybridization, the PSO-TS hybrids returned good solutions to the benchmark problems presented in the study. The hybrids also performed relatively fast. Furthermore, in some of the IP problems presented, the hybrids had shown better average solutions than pure PSO