Surveying the spirit of absolute summability on multilinear operators and homogeneous polynomials (Q259780): Difference between revisions

A linear operator \(u\in\mathcal L(E;F)\) is called absolutely \(p\)-summing if \(u\) sends weakly \(p\)-summable sequences of \(E\) to strongly \(p\)-summable sequences of \(F\). Equivalently, there is a constant \(C>0\) such that \[ \left(\sum_{j=1}^n\| u(x_j)\|^p\right)^{1/p}\leq C \sup_{x^* \in B_{E^*}} \left(\sum_{j=1}^n |x^*(x_j)|^p\right)^{1/p} \] for all \(n\) and for all \(x_1, \dots, x_n \in E\). This concept is essentially due to \textit{A. Grothendieck} and arises in his celebrated Résumeé [Bol. Soc. Mat. Sao Paulo 8, 1--79 (1956; Zbl 0074.32303)]. During the last five decades, many researchers have worked on this class of operators and a lot of interesting results and applications have been presented. We refer to the excellent book [\textit{J. Diestel} et al., Absolutely summing operators. Cambridge: Cambridge Univ. Press (1995; Zbl 0855.47016)] that contains a large variety of results related to this class. In the 1970's, \textit{A. Pietsch} systematized in an abstract way the classes of operators that improve the summability of vector-valued sequences, like the absolutely summing operators, defining the concept of operator ideals (see his monograph [Operator ideals. Amsterdam-New York-Oxford: North-Holland Publishing Company (1980; Zbl 0434.47030)]) and in the 1980's, he established the firsts steps to lead this abstract environment to the nonlinear framework (see [Forschungsergeb., Friedrich-Schiller-Univ. Jena 53, 16 p. (1983; Zbl 0561.47037)]). It is known that there are several extensions for the concept of absolutely summing linear operators to the multilinear and polynomial contexts (cf. [\textit{E. Çalışkan} and \textit{D. M. Pellegrino}, Rocky Mt. J. Math. 37, No. 4, 1137--1154 (2007; Zbl 1152.46034)]) and an important issue is to compare, by some means, these generalizations. This interesting paper is devoted to make an overall view of the known nonlinear generalizations of absolutely summing operators, comparing them from the point of view of which linear results are shared by each class in the multilinear context. Examples of results taken into account in this comparison are the Dvoretzky-Rogers theorem, Grothendieck's theorem, the Pietsch domination theorem, inclusion and factorization theorems. All presented extensions, in greater or lesser extent, do not have ``a good behavior'' and do not meet all expected properties. So the authors introduce a new generalization that fits all these (presented) linear properties to the multilinear context.