程序中采用了“共轭梯度——最小二乘”优化方法，其收敛性能较好，在较坏的初值情况下也能较快收敛。
In this program, the optimization method of "Conjugate Gradient-Least Squares" is employed, since it usually gives good convergent results.
应用了三种程序，它们分别是随机梯度(SG)、快速卡尔曼滤波(FK)和递归最小二乘(rls)。
Three programs, i. e., stochastic gradient (SG), fast Kalman (FK) and recursive least square (RLS), are used.
比较传统的最小二乘反演方法，利用共轭梯度的电阻率三维最小构造反演显示了更好的效果。
The 3-d resistivity inversion for minimum stricture using conjugate gradients shows good results as compared with traditional inversion method of least squares.
应用推荐