关于牛顿-拉弗森方法的一些应用
On Some Applications to the Newton--Raphson Method
AI总结:
本文从求积视角分析牛顿迭代,建立积分精确分解并给出误差显式界,连接牛顿法与复合梯形法则。
摘要:
牛顿方法在经典上被视为逼近非线性函数零点的迭代过程。在本文中,我们从求积的角度考虑牛顿迭代:迭代生成初始点与根之间区间的一个划分,这进而诱导出一个复合梯形法则。我们建立了与该牛顿生成求积相关的积分的精确分解,并刻画了由此产生的求积误差。我们还以初始牛顿衰减量给出了该误差的显式界,并给出了一个条件,在该条件下该误差小于原始区间上梯形法则的误差。这些结果提供了牛顿方法与其迭代诱导的求积法则之间的联系。
英文摘要:
Newton's method is classically viewed as an iterative procedure for approximating a zero of a nonlinear function. In this paper, we consider the Newton iterates from a quadrature perspective: the iterates generate a partition of the interval between the initial point and the root, which in turn induces a composite trapezoidal rule. We establish an exact decomposition of the integral associated with this Newton-generated quadrature and characterize the resulting quadrature error. We also obtain an explicit bound for this error in terms of the initial Newton decrement and give a condition under which it is smaller than the error of the trapezoidal rule on the original interval. These results provide a connection between Newton's method and the quadrature rule induced by its iterates.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。