BFGS

  拟牛顿法和最速下降法(Steepest Descent Methods)一样只要求每一步迭代时知道目标函数的梯度。通过测量梯度的变化，构造一个目标函数的模型使之足以产生超线性收敛性。这类方法大大优于最速下降法，尤其对于困难的问题。另外，因为拟牛顿法不需要二阶导数的信息，所以有时比牛顿法(Newton s Method)更为有效。如今，优化软件中包含了大量的拟牛顿算法用来解决无约束，约束，和大规模的优化问题。(The quasi-Newton method and the Steepest Descent Methods only require that each step iterations know the gradient of the objective function. By measuring the change of the gradient, constructing a model of the objective function is sufficient to produce superlinear convergence. This method is much better than the steepest descent method, especially for difficult problems. In addition, because the quasi-Newton method does not require information on the second derivative, it is sometimes more effective than the Newton s Method. Today, the optimization software contains a large number of quasi-Newton algorithm used to solve the unconstrained, constraint, and large-scale optimization problems.)

BFGS.m,526,2016-05-18
ex03_13062324.m,750,2016-05-18
Golden_section.m,749,2016-05-18
grad.m,852,2016-05-18