Is transpose of a matrix pseudo-inverse?

310 Views Asked by At

I think that if $A\vec{x} = \vec{b}$, then $A^{T}\vec{b} = \lambda \vec{x}$. I think this because any matrix $A$ can be decomposed into

$$ A = rotation1 \leftarrow scaling \leftarrow rotation2$$

and $A^{-1}$ I think should be

$$ A^{-1} = rotation2^{-1}\leftarrow scaling^{-1}\leftarrow rotation1^{-1}$$

and $A^{T}$ should also be very similar

$$A^{T} = rotation2^{-1} \leftarrow scaling \leftarrow rotation1^{-1}$$

Now, the transpose and the inverse only differ in scale. So I think that $$ if\quad A^{-1}\vec{b} = \vec{x}, \quad\quad then \quad A^{T}\vec{b} = \lambda\vec{x}$$ Now, I have just started learning about the four fundamental spaces of a matrix, and I am confused about what should be the geometric meaning of $A^{T}$. My intuition is made up of the following image.Diagram of the fundamental spaces of a matrix If you could tell me where my intuition is wrong in the above equations and the essence of the transpose of a matrix, then that will be of great help. Thanks...