Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
8896810 | Journal of Functional Analysis | 2018 | 31 Pages |
Abstract
Let n be a sufficiently large natural number and let B be an origin-symmetric convex body in Rnin the â-position, and such that the space (Rn,ââ
âB) admits a 1-unconditional basis. Then for any εâ(0,1/2], and for random cεlogâ¡n/logâ¡1ε-dimensional subspace E distributed according to the rotation-invariant (Haar) measure, the section Bâ©E is (1+ε)-Euclidean with probability close to one. This shows that the “worst-case” dependence on ε in the randomized Dvoretzky theorem in the â-position is significantly better than in John's position. It is a previously unexplored feature, which has strong connections with the concept of superconcentration introduced by S. Chatterjee. In fact, our main result follows from the next theorem: Let B be as before and assume additionally that B has a smooth boundary and Eγnââ
âBâ¤ncEγnâgradB(â
)â2 for a small universal constant c>0, where gradB(â
) is the gradient of ââ
âB and γn is the standard Gaussian measure in Rn. Then for any pâ[1,clogâ¡n] the p-th power of the norm ââ
âBp is Clogâ¡n-superconcentrated in the Gauss space.
Keywords
Related Topics
Physical Sciences and Engineering
Mathematics
Algebra and Number Theory
Authors
Konstantin Tikhomirov,