电子学报2001，Vol.29Issue(1)：61-63,3.
几种优化方法在脑磁逆问题中的应用与比较
Application and Comparison of Some Optimization Methods on MEG Inverse Problem
摘要
Abstract
Magnetic source parameter estimation by MEG data is a basic question on MEG research. Optimization methods are available to the estimation. Gradient, Gauss-Newton and simulated annealing algorithms are used separately. Computer simulation demonstrates that these algorithms are all availahle under certain conditions. In comparison, gradient, Gauss-Newton algorithms are of fast computation speed while simulated annealing costs more time. On the other hand, simulated annealing almost has no special need for the selection of initial iterative values while gradient and Gauss-Newton algorithms have such need to some extent. Furthermore,on estimation of multiple source parameters,gradient, Gauss-Newton algorithms sometimes fail due to incorrect selection of initial iterative values, while simulated annealing algorithm is still effective.关键词
脑磁图/逆问题/优化方法/梯度法/高斯-牛顿法/模拟退火法分类
生物科学引用本文复制引用
李军..几种优化方法在脑磁逆问题中的应用与比较[J].电子学报,2001,29(1):61-63,3.
