Generalized Simulated Annealing
Generalized Simulated Annealing
复制标题
DOI：
10.5772/66071
复制
发表时间：
2017-04
期刊：
影响因子：
--
通讯作者：
Y. Xiang;Sylvain Gubian;Florian Martin
中科院分区：
文献类型：
--
作者：
Y. Xiang;Sylvain Gubian;Florian Martin
数学、统计学、金融学、生物学、药理学、物理学、应用数学、经济学和化学中的许多问题都涉及到多维实值函数的全局最小值的确定。模拟退火算法已被广泛应用于不同的全局优化问题。已经开发了多种版本的模拟退火，包括经典模拟退火（CSA）、快速模拟退火（FSA）和广义模拟退火（GSA）。在使用Tsallis统计重新审视了GSA的基本思想之后，我们使用R包GenSA实现了一种改进的GSA方法。该软件包旨在解决具有大量局部极小值的复杂非线性目标函数。在本章中，我们将简要介绍这个R包，并通过解决不同领域的非凸优化问题来展示它的实用性：物理学，环境科学和金融。我们对GenSA与其他广泛使用的R包（包括rgenoud和DEoptim）进行了全面比较。GenSA非常有用，可以提供与其他广泛使用的R软件包相媲美甚至更好的优化解决方案。
Many problems in mathematics, statistics, finance, biology, pharmacology, physics, applied mathematics, economics, and chemistry involve the determination of the global minimum of multidimensional real-valued functions. Simulated annealing methods have been widely used for different global optimization problems. Multiple versions of simulated annealing have been developed, including classical simulated annealing (CSA), fast simulated annealing (FSA), and generalized simulated annealing (GSA). After revisiting the basic idea of GSA using Tsallis statistics, we implemented a modified GSA approach using the R package GenSA. This package was designed to solve complicated nonlinear objective functions with a large number of local minima. In this chapter, we provide a brief introduction to this R package and demonstrate its utility by solving non-convexoptimization problems in different fields: physics, environmental science, and finance. We performed a comprehensive comparison between GenSA and other widely used R packages, including rgenoud and DEoptim. GenSA is useful and can provide a solution that is comparable with or even better than that provided by other widely used R packages for optimization.