Three Binary Versions of Tree-Seed Algorithm for Binary Optimization

This study explores the adaptation of the Tree-Seed Algorithm (TSA), a population-based optimization method, for solving binary optimization problems. Initially designed for problems featuring continuous solution spaces, TSA is adjusted to address binary-structured optimization challenges. Three distinct approaches, namely the sigmoid function, modulo function, and xor logic gate, are employed to address TSA for binary optimization problem-solving. The efficacy of these methods is evaluated through experimentation on Uncapacitated Facility Location Problems (UFLPs), representative pure binary problems from existing literature. A comprehensive analysis is conducted using a selection of well-known small, medium, and large-sized UFLP instances to assess the performances and the impact of TSA's control parameters. Comparative analysis of obtained results reveals promising outcomes achieved by the proposed algorithm.