Low rank matrix and compressed sensing

This is a note for part III of Linear Algebra and learning from data, Gilbert Strang

The main themes are sparsity (Low rank), Information theory (compression), and of course linear transformation.

A full rank matrix is inefficient. Finding low lank matrix which is close with original matrix can save computation.

The rank one matrix \(uv^{T}\) is a unit of a matrix. The full rank matrix can be decomposed by sum of rank one matrices i.e. singular vector decomposition.

Sherman–Morrison formula suggests update rule for adding rank one matrix to the original matrix.

\[(A + \mathbf{u} \mathbf{v}^{T})^{-1} = A^{-1} - \frac{A^{-1} \mathbf{u} \mathbf{v}^{T}A^{-1}}{1 + \mathbf{v}^{T} A^{-1} \mathbf{u}}\]

The matrix norm is associated with singular value, \(\sigma\).

The point is the unit of matrix is the rank one matrix, specially outer product of singular vectors \(uv^{T}\). \(uv^{T}\) is a coordinate of the matrix space and singular value \(\sigma\) is a point on the coordinate.

System, Inner product, \(A^{T}A\), Steady state equilibrium, dual

A system has a law. The observations follow the law. The state set by the system’s law. The state has two variables.

\[ Ax=b \;(1)\\ Y= \beta X \; (2)\\ \hat{x} = \frac {A^{T}b}{(A^{T}A) \; (3)\]

Equation (1) is observations, (2) is a system, (3) is fit the observations to the system. \(A^{T}A\) is a steady state equilibrium. \(A\) and \(A^{T}\) are dual. \(x\) and \(b\) are dual.

Jun Kang
Clinical Assistant Professor of Hospital Pathology

My research interests include pathology, oncology and statistics.