K.Yu. Osipenko
Abstract:The paper concerns problems of the recovery of operators from noisy information in weighted $L_q$-spaces with homogeneous weights. A number of general theorems are proved and applied to finding exact constants in multidimensional Carlson type inequalities with several weights and problems of the recovery of differential operators from a noisy Fourier transform. In particular, optimal methods are obtained for the recovery of powers of generalized Laplace operators from a noisy Fourier transform in the $L_p$-metric.
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is the problem of recovering operators from noisy information in weighted \(L^q\) spaces, especially for weights with symmetric properties. Specifically, the article mainly focuses on the following aspects:
1. **Optimization recovery problem**: Research how to optimally recover the power form of the generalized Laplacian operator from noisy Fourier transform information. This involves the optimal recovery method in the \(L^p\) metric.
2. **Generalized Carlson inequality**: Look for the exact constants of multi - dimensional Carlson - type inequalities involving multiple weights. For example, for an inequality of the form
\[
\left\| w(\cdot)x(\cdot) \right\|_{L^q(T,\mu)} \leq K \left\| w_0(\cdot)x(\cdot) \right\|^\gamma_{L^p(T,\mu)} \left( \max_{1 \leq j \leq n} \left\| \omega_j(\cdot)x(\cdot) \right\|_{L^r(T,\mu)} \right)^{1-\gamma}
\]
find the optimal constant \(K\).
3. **Extremum problem**: Consider the following extremum problem:
\[
\left\| w(\cdot)x(\cdot) \right\|_{L^q(T,\mu)} \to \max, \quad \left\| w_0(\cdot)x(\cdot) \right\|_{L^p(T,\mu)} \leq \delta, \quad \left\| w_j(\cdot)x(\cdot) \right\|_{L^r(T,\mu)} \leq 1, \quad j = 1,\ldots,n,
\]
where \(w(\cdot)\), \(w_0(\cdot)\), and \(w_j(\cdot)\) are homogeneous functions with certain symmetric properties.
4. **Optimal recovery error**: Define and calculate the optimal recovery error \(E(p,q,r)\), that is, the smallest error among all possible recovery methods under given conditions. The specific formula is:
\[
E(p,q,r) = \inf_{m: L^p(T_0,\mu) \to L^q(T,\mu)} e(p,q,r,m),
\]
where \(e(p,q,r,m)\) represents the error of method \(m\).
By solving these problems, the paper aims to provide a theoretical framework and specific methods to deal with the problem of recovering operators from noisy data, and give relevant optimal constants and inequalities.