首先用构造性的方法证明:对于任意的 n 阶多元多项式函数,可以构造一个三层前向神经网络以任意精度逼近该多项式,所构造网络的隐层节点个数仅与多项式的维数 d 和阶数 n 有关.然后,我们给出实现这一逼近的具体算法.最后,给出两个算例进...首先用构造性的方法证明:对于任意的 n 阶多元多项式函数,可以构造一个三层前向神经网络以任意精度逼近该多项式,所构造网络的隐层节点个数仅与多项式的维数 d 和阶数 n 有关.然后,我们给出实现这一逼近的具体算法.最后,给出两个算例进一步验证所得的理论结果.本文结果对神经网络逼近多元多项式函数的具体网络构造以及实现这一逼近的方法等问题具有指导意义.展开更多
A new necessary and sufficient condition for the existence of minor left prime factorizations of multivariate polynomial matrices without full row rank is presented.The key idea is to establish a relationship between ...A new necessary and sufficient condition for the existence of minor left prime factorizations of multivariate polynomial matrices without full row rank is presented.The key idea is to establish a relationship between a matrix and any of its full row rank submatrices.Based on the new result,the authors propose an algorithm for factorizing matrices and have implemented it on the computer algebra system Maple.Two examples are given to illustrate the effectiveness of the algorithm,and experimental data shows that the algorithm is efficient.展开更多
In this paper, we have obtained an expression of the bivariate Vandermonde determinant for the Elliptic Type Node Configuration in R-2, and discussed the possibility of the corresponding multivariate Lagrange, Hermite...In this paper, we have obtained an expression of the bivariate Vandermonde determinant for the Elliptic Type Node Configuration in R-2, and discussed the possibility of the corresponding multivariate Lagrange, Hermite and Birkhoff interpolation.展开更多
In this study, a multivariate local quadratic polynomial regression(MLQPR) method is proposed to design a model for the sludge volume index(SVI). In MLQPR, a quadratic polynomial regression function is established to ...In this study, a multivariate local quadratic polynomial regression(MLQPR) method is proposed to design a model for the sludge volume index(SVI). In MLQPR, a quadratic polynomial regression function is established to describe the relationship between SVI and the relative variables, and the important terms of the quadratic polynomial regression function are determined by the significant test of the corresponding coefficients. Moreover, a local estimation method is introduced to adjust the weights of the quadratic polynomial regression function to improve the model accuracy. Finally, the proposed method is applied to predict the SVI values in a real wastewater treatment process(WWTP). The experimental results demonstrate that the proposed MLQPR method has faster testing speed and more accurate results than some existing methods.展开更多
This paper presents a new algorithm for computing the extended Hensel construction(EHC) of multivariate polynomials in main variable x and sub-variables u1, u2, ···, um over a number field K. This algor...This paper presents a new algorithm for computing the extended Hensel construction(EHC) of multivariate polynomials in main variable x and sub-variables u1, u2, ···, um over a number field K. This algorithm first constructs a set by using the resultant of two initial coprime factors w.r.t. x, and then obtains the Hensel factors by comparing the coefficients of xi on both sides of an equation. Since the Hensel factors are polynomials of the main variable with coefficients in fraction field K(u1, u2, ···, um), the computation cost of handling rational functions can be high. Therefore,the authors use a method which multiplies resultant and removes the denominators of the rational functions. Unlike previously-developed algorithms that use interpolation functions or Grobner basis, the algorithm relies little on polynomial division, and avoids multiplying by different factors when removing the denominators of Hensel factors. All algorithms are implemented using Magma, a computational algebra system and experiments indicate that our algorithm is more efficient.展开更多
In this paper, we consider the higher divided difference of a composite function f（g（t）） in which g（t） is an s-dimensional vector. By exploiting some properties from mixed partial divided differences and multiva...In this paper, we consider the higher divided difference of a composite function f（g（t）） in which g（t） is an s-dimensional vector. By exploiting some properties from mixed partial divided differences and multivariate Newton interpolation, we generalize the divided difference form of Faà di Bruno＇s formula with a scalar argument. Moreover, a generalized Faà di Bruno＇s formula with a vector argument is derived.展开更多
The use of Bernstein-Bézier net in the study of bivariate splines was initiated by, G. Farin. In [1], Farin used Bèzier coordinates to express C^r continuity condition for bivariate splines. In [2], de Boor ...The use of Bernstein-Bézier net in the study of bivariate splines was initiated by, G. Farin. In [1], Farin used Bèzier coordinates to express C^r continuity condition for bivariate splines. In [2], de Boor and H(?)llig applied B-net method to obtain the approximation order of the space of C^1-cubic bivariate splines on a three-directionmesh. In this note, we study the B-net representation of multivariate splines. In展开更多
In this paper, we study the relationship between iterated resultant and multivariate discriminant. We show that, for generic form f（xn） with even degree d, if the polynomial is squarefreed after each iteration, the ...In this paper, we study the relationship between iterated resultant and multivariate discriminant. We show that, for generic form f（xn） with even degree d, if the polynomial is squarefreed after each iteration, the multivariate discriminant A（f） is a factor of the squarefreed iterated resulrant. In fact, we find a factor Hp（f, [x1 , xn]） of the squarefreed iterated resultant, and prove that the multivariate discriminant A（f） is a factor of Hp（f,[x1,... ,xn]）. Moreover, we conjecture that Hp（f, [x1,..., xn]） =△（f） holds for generic form f, and show that it is true for generic trivariate form f（x, y, z）.展开更多
文摘首先用构造性的方法证明:对于任意的 n 阶多元多项式函数,可以构造一个三层前向神经网络以任意精度逼近该多项式,所构造网络的隐层节点个数仅与多项式的维数 d 和阶数 n 有关.然后,我们给出实现这一逼近的具体算法.最后,给出两个算例进一步验证所得的理论结果.本文结果对神经网络逼近多元多项式函数的具体网络构造以及实现这一逼近的方法等问题具有指导意义.
基金supported by the National Natural Science Foundation of China under Grant Nos.12171469,12001030 and 12201210the National Key Research and Development Program under Grant No.2020YFA0712300the Fundamental Research Funds for the Central Universities under Grant No.2682022CX048。
文摘A new necessary and sufficient condition for the existence of minor left prime factorizations of multivariate polynomial matrices without full row rank is presented.The key idea is to establish a relationship between a matrix and any of its full row rank submatrices.Based on the new result,the authors propose an algorithm for factorizing matrices and have implemented it on the computer algebra system Maple.Two examples are given to illustrate the effectiveness of the algorithm,and experimental data shows that the algorithm is efficient.
文摘In this paper, we have obtained an expression of the bivariate Vandermonde determinant for the Elliptic Type Node Configuration in R-2, and discussed the possibility of the corresponding multivariate Lagrange, Hermite and Birkhoff interpolation.
文摘In this study, a multivariate local quadratic polynomial regression(MLQPR) method is proposed to design a model for the sludge volume index(SVI). In MLQPR, a quadratic polynomial regression function is established to describe the relationship between SVI and the relative variables, and the important terms of the quadratic polynomial regression function are determined by the significant test of the corresponding coefficients. Moreover, a local estimation method is introduced to adjust the weights of the quadratic polynomial regression function to improve the model accuracy. Finally, the proposed method is applied to predict the SVI values in a real wastewater treatment process(WWTP). The experimental results demonstrate that the proposed MLQPR method has faster testing speed and more accurate results than some existing methods.
基金supported in part by the National Natural Science Foundation of China under Grant No.11371356CAS Project QYZDJ-SSW-SYS022the Strategy Cooperation Project AQ-1701
文摘This paper presents a new algorithm for computing the extended Hensel construction(EHC) of multivariate polynomials in main variable x and sub-variables u1, u2, ···, um over a number field K. This algorithm first constructs a set by using the resultant of two initial coprime factors w.r.t. x, and then obtains the Hensel factors by comparing the coefficients of xi on both sides of an equation. Since the Hensel factors are polynomials of the main variable with coefficients in fraction field K(u1, u2, ···, um), the computation cost of handling rational functions can be high. Therefore,the authors use a method which multiplies resultant and removes the denominators of the rational functions. Unlike previously-developed algorithms that use interpolation functions or Grobner basis, the algorithm relies little on polynomial division, and avoids multiplying by different factors when removing the denominators of Hensel factors. All algorithms are implemented using Magma, a computational algebra system and experiments indicate that our algorithm is more efficient.
基金Acknowledgments. This work was supported by the National Science Foundation of China （Grant Nos. 10471128, 10731060）.
文摘In this paper, we consider the higher divided difference of a composite function f（g（t）） in which g（t） is an s-dimensional vector. By exploiting some properties from mixed partial divided differences and multivariate Newton interpolation, we generalize the divided difference form of Faà di Bruno＇s formula with a scalar argument. Moreover, a generalized Faà di Bruno＇s formula with a vector argument is derived.
文摘The use of Bernstein-Bézier net in the study of bivariate splines was initiated by, G. Farin. In [1], Farin used Bèzier coordinates to express C^r continuity condition for bivariate splines. In [2], de Boor and H(?)llig applied B-net method to obtain the approximation order of the space of C^1-cubic bivariate splines on a three-directionmesh. In this note, we study the B-net representation of multivariate splines. In
文摘In this paper, we study the relationship between iterated resultant and multivariate discriminant. We show that, for generic form f（xn） with even degree d, if the polynomial is squarefreed after each iteration, the multivariate discriminant A（f） is a factor of the squarefreed iterated resulrant. In fact, we find a factor Hp（f, [x1 , xn]） of the squarefreed iterated resultant, and prove that the multivariate discriminant A（f） is a factor of Hp（f,[x1,... ,xn]）. Moreover, we conjecture that Hp（f, [x1,..., xn]） =△（f） holds for generic form f, and show that it is true for generic trivariate form f（x, y, z）.
