Norm Estimates for the Difference Between Bochner's Integral and the Convex Combination of Function's Values
P. Cerone, Y. J. Cho, S. S. Dragomir, J. K. Kim, S. S. Kim
TL;DR
The paper addresses estimating the error when approximating the Bochner integral of a vector-valued function by a convex combination of its values at prescribed nodes in a Banach space with the Radon-Nikodym property. It develops a kernel-based integral identity and derives Ostrowski-type bounds for vector-valued functions, with explicit dependence on $f'$, the auxiliary quantities $\mu_p$, and the chosen $L_p$ norms. The authors specialize the general results to two- and three-point rules, including trapezoidal and Simpson-type configurations, and discuss optimal node placements and weight choices to minimize the error. Overall, the work extends scalar Ostrowski inequalities to the vector-valued, Banach-space setting and provides practical, norm-based error estimates for quadrature of Banach-space-valued functions.
Abstract
Norm estimates are developed between the Bochner integral of a vector-valued function in Banach spaces having the Radon-Nikodym property and the convex combination of function values taken on a division of the interval [a,b].
