高阶方向导数及其应用

    An Extension to High Order for Directional Derivative of Multivariate Function

    • 摘要: 将多元函数方向导数概念予以推广,在得到二阶方向导数定义和计算公式后,给出了多元函数的高阶方向导数.提出了高阶方向导数的应用:1)把一元函数性质推广到多元函数的一般途径;2)得到多元函数取极值的必要条件和充分必要条件;3)利用二阶方向导数解释了矩阵半正定和半负定的几何意义;4)揭示出线性方程组当矩阵正定或负定时,背后存在的一个极值问题.5)推导出多元函数的Taylor展式.

       

      Abstract: The directional derivative concept of the multivariate function is expanded from first order to high order in this paper. After getting the definition of second order directional derivative and the calculation formula,the high order directional derivative has been given. Applications of the high order directional derivative are proposed as following: 1) A general way to expanding simple variable function characteristics to multivariate function is presented. 2) The necessary conditions,necessary and sufficient conditions of extreme value criterion of function are easily obtained. 3) The geometrical meaning of semi-positive and semi-negative definite is explained according to second-order directional derivatives. 4) It is revealed that there is extreme value problem of the function when the matrix of the linear equations is positive or negative definite. 5) Taylor's expansion formula of the multivariate function is easily deduced.

       

    /

    返回文章
    返回