Interdisciplinary Program of Bioengineering, Seoul National University, 599 Gwanangno, Gwanak-Gu, Seoul, 151-742, Korea.
Lasers Med Sci. 2010 Sep;25(5):733-41. doi: 10.1007/s10103-010-0793-x. Epub 2010 Jun 12.
A hybrid deterministic-stochastic algorithm combining the simplex method (SM) and a genetic algorithm (GA) was applied to the problem of extracting the optical and morphological properties of human skin (HSOMPs) from visual reflectance spectroscopy data. The results using the GA-SM hybrid algorithm adopting tournament selection and selecting new sets of HSOMPs were compared with those using other conventional optimization algorithms that have generally been used for the extraction of HSOMPs. Monte-Carlo simulation showed that the suggested GA-SM hybrid algorithm enhanced the stability of the inverse solutions and computational efficiency.
一种将单纯形法（SM）和遗传算法（GA）相结合的混合确定性-随机性算法被应用于从视觉反射光谱数据中提取人体皮肤的光学和形态学特性（HSOMPs）的问题。使用遗传算法-SM 混合算法（采用竞赛选择并选择新的 HSOMP 集）的结果与通常用于提取 HSOMPs 的其他常规优化算法的结果进行了比较。蒙特卡罗模拟表明，所提出的 GA-SM 混合算法提高了逆解的稳定性和计算效率。
