【Numberical Optimization】5 Conjugate Gradient Methods (zen学习笔记)
5.1 The linear conjugate gradient methods
The conjugate gradient method is an itterative method for solving a linear system of equations:
which can be stated equivalently as the following minimization problem:
第k步的残差:
5.1.1 Conjugate direction methods
1.阐述了共轭方向算法能在有限步内收敛到最优点(步数主要与特征值分布有关)
The conjugate vectors are given.
-
is said to be conjugate with respect to the symmetric positive definite martix A if
-
is also linearly independent.
- 有限步内能收敛到最优解,证明及算法如下:
Defining as
where S is defined by
then
2.最优点在集合中 即可由
线性表出
5.1.2 Basic properties of the conjugate gradient method
A PRACTICAL FORM OF THE CONJUGATE GRADIENT METHOD
前面介绍了共轭梯度算法的一些性质,接下来的算法提出共轭方向的构造
共轭梯度初步算法及共轭梯度算法
改进之处在于:
-
5.1.3 RATE OF CONVERGENCE
解释收敛步数与特征值分布的关系、验证 CG算法在某种意义上是最优的