Newton Cotes methods for integration of fuzzy functions (Q2485634)

The paper deals with the numerical integration of fuzzy functions by Newton-Cotes formulas. It is not quite clear how the author defines a ``Newton-Cotes formula''. Up to Theorem 3.2 he seems to follow the standard convention (i.e. an interpolatory quadrature formula with \(n\) equidistant nodes); in the later part of the paper he then switches to an \(n\)-fold compound formula based on a low-order interpolatory formula (e.g. a trapezoidal or Simpson method). In Theorem 3.2 the classical error representation via the Peano kernel is carried over to the fuzzy setting. Theorem 3.1 gives a vaguely formulated ``convergence result''. Assuming that the author wants to investigate convergence as \(n\to\infty\) (\(n\) being the number of quadrature nodes), as the notation suggests, the result is wrong because it is already incorrect in the classical (non-fuzzy) setting which can be considered as a special case of the fuzzy problem. The erroneous assumption in the ``proof'' is that the formulas (3.5) can be represented in the form indicated in Definition 2.3: A necessary (but not sufficient) condition for this to be true would be that all weights of the quadrature are positive, but this is known to be violated for the Newton-Cotes formulas.