Lusin's Theorem and Bochner Integration

Peter A. Loeb,Erik Talvila
DOI: https://doi.org/10.48550/arXiv.math/0406370
2004-06-18
Classical Analysis and ODEs
Abstract:It is shown that the approximating functions used to define the Bochner integral can be formed using geometrically nice sets, such as balls, from a differentiation basis. Moreover, every appropriate sum of this form will be within a preassigned $\epsilon$ of the integral, with the sum for the local errors also less than $\epsilon$. All of this follows from the ubiquity of Lebesgue points, which is a consequence of Lusin's theorem, for which a simple proof is included in the discussion.
What problem does this paper attempt to address?