Eigenimages Singular Value Decomposition (SVD)
Eigenimages Singular Value Decomposition (SVD)
- SVD is a mathematical technique which can express any
matrix P as the product of three matrices.
- SVD requires that the number of rows, r, in P be
equal to or greater than than the number of
columns, c.
- If this is not the case, there are
“tricks” one can apply to get around
this constraint.
- svd{P} = T*S*E where T*S*ET = P.
- T is an r*c column-orthogonal matrix.
- S is and c*c diagonal matrix with diagonal elements >=
0.
- E is an c*c row and column orthogonal matrix.