• Medientyp: E-Book
  • Titel: Black box low tensor rank approximation using fibre-crosses
  • Beteiligte: Espig, Mike [Verfasser:in]; Grasedyck, Lars [Verfasser:in]; Hackbusch, Wolfgang [Verfasser:in]
  • Erschienen: Leipzig: Max-Planck-Inst. f. Mathematik in den Naturwiss., 2008
  • Erschienen in: Max-Planck-Institut für Mathematik in den Naturwissenschaften: Preprints ; 2008060
  • Ausgabe: rev. version: May 2009
  • Umfang: Online-Ressource (38 S., 395 KB)
  • Sprache: Englisch
  • Entstehung:
  • Anmerkungen:
  • Beschreibung: In this article we introduce a black box type approximation algorithm for tensors A in high dimension d. The algorithm determines adaptively the positions of entries of the tensor that have to be computed or read, and using these (few) entries it constructs a low rank tensor approximation X that minimizes the l2-distance between A and X at the chosen positions. The full tensor A is not required, only the evaluation of A at a few positions. The minimization problem is solved by Newton's method which requires the computation and evaluation of the Hessian. For efficiency reasons the positions are located on fibre-crosses of the tensor so that the Hessian can be assembled and evaluated in a data-sparse form requiring a complexity of O(Pd), where P is the number of fibre-crosses and d the order of the tensor.
  • Zugangsstatus: Freier Zugang