A High-Dimensional Particle Swarm Optimization Based on Similarity Measurement

摘要

Particle Swarm Optimization (PSO) is a kind of classical population-based intelligent optimization methods that widely used in solving various optimization problems. With the increase of the dimensions of the optimized problem, the high-dimensional particle swarm optimization becomes an urgent, practical and popular issue. Based on data similarly measurement, a high-dimensional PSO algorithm is proposed to solve the high-dimensional problems. The study primarily defines a new distance paradigm based on the existing similarity measurement of high-dimensional data. This is followed by proposes a PSO variant under the new distance paradigm, namely the LPSO algorithm, which is extended from the classical Euclidean space to the metric space. Finally, it is showed that LPSO could obtain better solution at higher convergence speed in high-dimensional search space.