Lp approximation of variational problems in L1 and L∞ (Q4239805)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 1279366
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Lp approximation of variational problems in L1 and L∞ |
scientific article; zbMATH DE number 1279366 |
Statements
Lp approximation of variational problems in L1 and L∞ (English)
0 references
4 April 2000
0 references
\(L^p\) approximation
0 references
\(L^\infty\) constraints
0 references
hierarchical minimization
0 references
Tchebyshev approximation
0 references
natural median
0 references
semi-coercive variational inequalities
0 references
0.7861961126327515
0 references
0.7799340486526489
0 references
This very consistent paper contains a large number of results (some 19 theorems, propositions, lemmas and corollaries), concerning asymptotic properties, either as \(p \to \infty\) or as \(p \to 1\), of the unique solutions, \(u(p)\in X_p, \;p\in (1,\infty),\) of strictly convex variational problems of the form: \[ \min \{J(u)+\tfrac 1p(\|p\|_p)^p; \;u\in X_p \} \tag \(P_p\) \] defined by closed proper convex functionals \(J(.):X_p \to R\cup \{\infty\}\). In the finite dimensional case, \(X_p=R^n \;\forall p\in [0,\infty]\), the main result (Theorem 2.1) is relatively simple and states, essentially, that the set \(\{u(p), \;p\in (1,\infty) \}\subset R^n \) is bounded as \(p\to \infty\) and every accumulation point \(u_\infty =\lim_{p_k \to \infty}u(p_k)\) is a solution of the \(L^\infty\) constrained problem: \[ \min \{J(u); \;\|u\|_\infty \leq 1 \} \tag \(P_\infty\) \] and solves also the problem: \[ \min \{\|u\|_\infty; \;u\in S(P_\infty) \} \tag \(P_\infty^1\) \] while if the problem \((P_1)\) has a nonempty set of solutions \(S(P_1)\), then \(\exists \;\lim_{p \to 1}u(p)=x_1\in S(P_1)\) and moreover, \(x_1\) is the unique solution of the associated problem: \[ \min \Biggl\{\sum_{i=1}^n |u_i|\ln(|u_i|/e); \;u\in S(P_1) \Biggr\}. \tag \(P_1^1\) \] In the infinite-dimensional case, in which \(X_p=L^p(\Omega)\), where \((\Omega,{\mathcal A},\mu)\) is a measure space with a positive finite measure, the analogous results are much more complicated. Some of these results are applied to certain semi-coercive variational inequalities, to the problem of natural median and, in the finite-dimensional case, to Tchebyshev approximations.
0 references