notes

Personal notes
git clone git://git.laack.co/notes.git
Log | Files | Refs

FrobeniusNorm.md (361B)


      1 # Frobenius Norm
      2 
      3 **Source:** Deep Learning
      4 
      5 **Chapter:** 2
      6 
      7 **Definition:** The Frobenius norm is a norm defined on matricies. This norm is defined as follows:
      8 
      9 ||A||_F = sqrt(sum(A^2 for all i,j))
     10 
     11 Basically, we sum each of the squared indices and then take the sqrt at the end. This is an analog of the L2 norm, but defined for matricies instead of vectors.