Weighted Reductive Multilinear Array Decomposition

Divanyan L., Demiralp M.

International Conference on Numerical Analysis and Applied Mathematics (ICNAAM), Halkidiki, Yunanistan, 19 - 25 Eylül 2011, cilt.1389 identifier identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası: 1389
  • Doi Numarası: 10.1063/1.3637820
  • Basıldığı Şehir: Halkidiki
  • Basıldığı Ülke: Yunanistan
  • İstanbul Teknik Üniversitesi Adresli: Evet

Özet

There are many possibilities to decompose a multilinear array (an array with more than two indices). Amongst which singular value decomposition, discrete high dimensional model representation, and, discrete enhanced multivariance product representation are most widely used ones. The difficulties of the singular value decomposition can be reduced to minimum by using reductive array decomposition (which was developed by Demiralps) where the ordinary linear algebraic tools can be directly used. This work extends the ordinary reductive array decomposition to its weighted version under diagonal weight matrices.