A comparative study of gradient-based and meta-heuristic optimization methods using Griewank benchmark function

Abstract

Optimization methods are frequently applied to solve real-world problems such, engineering design, computer science, and computational chemistry. This paper aims to compare gradient-based algorithms and the meta-heuristic particle swarm optimization to minimize the multidimensional benchmark Griewank function, a multimodal function with widespread local minima. Several approaches of gradient-based methods such as steepest descent, conjugate gradient with Fletcher-Reeves and Polak-Ribiere formulations, and quasi-Newton Davidon-Fletcher-Powell approach were compared. The results presented showed that the meta-heuristic method is recommended for function with this behavior because is no needed prior information of the search space. The performance comparison includes computation time and convergence of global and local optimum.

Ozonioterapia no tratamento de feridas em adultos: revisão integrativa / Ozoniotherapy in the treatment of wounds in adults: integrative review

“…De acordo com Rodrigues Júnior (2005), os métodos determinísticos, também chamados de métodos clássicos, compõem uma metodologia de otimização que envolve cálculo de derivadas de primeira ordem ou segunda ordem. Como alternativa aos métodos determinísticos, os métodos heurísticos, também chamados de probabilísticos, mitigam a necessidade de ter informações sobre derivadas e gradientes em seu procedimento de cálculo (Dalla et al, 2021), mas sim por encontrarem as soluções mais prováveis que otimizam a função objetivo iterativamente. Para tanto, os processos de otimização heurísticos são normalmente computacionalmente caros devido à elevada número de avaliações da função objetiva e de suas restrições para obter-se possíveis soluções para a função objetivo (Medeiros e Kripka, 2012).…”

Analysis of heat transmission in convective, radiative and moving rod with thermal conductivity using meta-heuristic-driven soft computing technique

“…In addition, since the 1950s, there has been substantial development of gradient-based techniques, and today there are multiple excellent options available for dealing with smooth nonlinear issues. These approaches only converge to a local minimum point for the objective function since the search process that they employ uses only local information (functions and their gradients at a place) (Dalla et al 2021). The gradient-based methods estimate the motion by analysis of the strong differences in brightness between analyzed regions.…”