Error level saturation for Popoff's generalized derivative operator
The generalized derivative of a function \(f\) introduced by K. Popoff is defined by \[ f^p(x)=\lim_{h\rightarrow 0} P_h f(x), \] provided that this limit exists, where \[ P_h f(x)= \frac{2}{h^2}\int_{0}^{h}[f(x+t)-f(x)]dt. \] Suppose that \(f\) has a continuous bounded second derivative on some open interval \(I\). For \(\varepsilon>0\), by \(f^\varepsilon\) is denoted a bounded integrable perturbation of \(f\) satisfying \(|f(t)-f^\varepsilon(t)|\leq\varepsilon\) for \(t\in I\). It is shown that \[ |P_{\sqrt{\varepsilon}}f^{\varepsilon}(x)-f^{\prime}(x)|=O(\sqrt{\varepsilon}); \] and on the other hand, the following saturation result is established: If for some choice \(h(\varepsilon)\rightarrow 0\), as \(\varepsilon\rightarrow 0\), \[ |P_{h(\varepsilon)}f^{\varepsilon}(x)-f^{\prime}(x)|=o(\sqrt{\varepsilon}) \] for each \(x\in I\), then \(f\) is a linear function on \(I\).
This page was built for publication: Error level saturation for Popoff's generalized derivative operator
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2392535)
