MUCPSO: A Modified Chaotic Particle Swarm Optimization with Uniform Initialization for Optimizing Software Effort Estimation

Particle Swarm Optimization is a metaheuristic optimization algorithm widely used across broad range of applications. The has certain primary advantages such as its ease implementation, high convergence accuracy, and fast speed. Nevertheless, since origin in 1995, swarm still suffers from two shortcomings, i.e., premature easy trapping local optima. Therefore, this study proposes modified chaotic particle with uniform initialization to enhance the comprehensive performance standard by introducing three additional schemes. Firstly, initialized generated through approach. Secondly, replacing linear inertia weight nonlinear map. Thirdly, applying personal learning strategy global search avoid trap proposed examined compared optimization, recent variants, nature-inspired using software effort estimation methods benchmark functions: Use case points, COCOMO, Agile. Detailed investigations prove that schemes work well develop an exploitative manner, which created avoids being trapped on optimum solution explorative manner chaotic-based weight.