scientific article; zbMATH DE number 1750316
Let \((G,+)\) be a uniquely \(n!\)-divisible abelian semigroup and \(H\) a group. A polynomial function (``function polynomial in the author's terminology) of degree \(n\) is a solution \(f: G\to H\) of the functional equation \(\Delta_h^{n+1}f(x)=0,\) where \(\Delta_h^1 f(x)=f(x+h)-f(x)\) and \(\Delta_h^{j+1}f(x)=\Delta_h^1[\Delta_h^j f(x)]\) \((j=1,2,\dots).\) The author offers results connecting two kinds of stability for this equation, the Hyers-Ulam stability (roughly: if \(\Delta_h^{n+1}g\) is close to 0 then there exixts a polynomial function of degree \(n\) close to \(g\)) and almost everywhere stability (if \(\Delta_h^{n+1}g(x)=0\) almost everywhere then \(g\) equals a polynomial function of degree \(n\) almost everywhere). The theorems and proofs are highly technical. The main new idea is to consider ``almost everywhere with respect to subsets of \(G\) belonging to a family \(\mathcal{I}\) that is translation-invariant (\(\{x\in G\mid a+x\in A\}\in\mathcal{I}\) if \(a\in G\), \(A\in\mathcal{I}\)) and \(k\)-proper (\(\{A_1\cup\dots \cup A_k\mid A_1,\dots ,A_k\in\mathcal{I}\}\notin\mathcal{I}\)). [The paper is written in rudimentary (e.g. ``\dots there exists a unique up to constant function polynomial \dots ) but understandable English. In the last line of Theorem 2, \(a(x)\) should be corrected to \(p(x)\)].
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4532971)