Mathematics · Statistics and Probability
Norms in product spaces which preserve approximation properties
Carlos Alberto Velasco Benítez, Manuel Fernández
See on the mapAbstract
Synopsis Let E 1 and E 2 be real normed linear spaces such that the dimension of any of them is at least 2. We prove that the norms in E 1 × E 2 which verify a simple property of monotonicity with regard to the initial norms in E 1 and E 2 are the only norms in E 1 × E 2 which preserve best linear approximations, in the sense that if y k ∊ L k is best approximation to x k from the linear subspace L k , ( k = 1,2), then ( y 1 , y 2 ) is best approximation to ( x 1 , x 2 ) from L 1 × L 2 .