Representation theorems for indefinite quadratic forms without spectral gap

From MaRDI portal



Abstract: The First and Second Representation Theorem for sign-indefinite quadratic forms are extended. We include new cases of unbounded forms associated with operators that do not necessarily have a spectral gap around zero. The kernel of the associated operators is determined for special cases. This extends results by Grubiv{s}i'c, Kostrykin, Makarov and Veseli'c in [Mathematika 59 (2013), 169--189].


Self-adjoint operators on a Hilbert space define symmetric sesquilinear forms, but not every such form arises in this way, nor is the correspondence one-to-one in general, when one goes beyond the bounded case which is ruled by Riesz's representation theorem. The author extends the first representation theorem for indefinite forms to cover new cases (Theorem 2.3), building on the work of \textit{L. Grubišić} et al. [Mathematika 59, No. 1, 169--189 (2013; Zbl 1272.47004)]. Spectral gap conditions on the representing operators are dropped, and new assumptions have to be imposed in view of Example 1. In the special case of non-negative forms with off-diagonal perturbations, a complete generalization of the results of [loc. cit.] is given (Theorem 2.7), together with an explicit description of the kernel (Theorem 2.13), extending the analysis of [\textit{V. Kostrykin} et al., Oper. Theory Adv. Appl. 149, 349--372 (2004; Zbl 1084.47002)]. In order for the second representation theorem to hold as well, equivalent and sufficient conditions for the domain stability condition are presented (Theorem 3.8 and Lemma 3.9), as natural extensions of those contained in [Zbl 1272.47004)].











This page was built for publication: Representation theorems for indefinite quadratic forms without spectral gap

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q747182)