Rules for Solving Unknown Matrix A in Ax=b Problem?

  • Thread starter Genxi
  • Start date
  • Tags
    Matrix
In summary, the conversation discusses methods for solving a system of linear equations involving a 3x3 matrix A, a 3x1 matrix x, and a 3x1 matrix b. Some suggested methods include Cramer's rule, Gauss elimination, LU-factorization, and matrix inversion. It is also noted that the system may be underdetermined, meaning there may be no unique solution or an infinite number of solutions.
  • #1
Genxi
2
0

Homework Statement



I am trying to solve for the A matrix (3 x 3). I know matrix x is (3 x 1) and matrix b is (3 x 1), how do I go about solving for matrix A?


The Attempt at a Solution



I have not attempted this as I don't know the rules to initiate this problem.

Please offer me some hints or rules I should know on how to solve this problem


Thanks,


Genxi
 
Physics news on Phys.org
  • #2
Well you could re-arrange it with matrix algebra:

[itex]Ax=b[/itex]

[itex]Axx^{T}=bx^{T}[/itex]

[itex]A=\frac{bx^{T}}{xx^{T}}[/itex]

Note the last step is allowed because [itex]xx^{T}[/itex] is a scalar.(You can't divide matrices by matrices)

Familiarise yourself with the transpose of a matrix (in this case, a column vector) and matrix multiplication, and perhaps rules of matrix algebra and that should be all you need to understand the above.

This link may be helpful http://people.hofstra.edu/stefan_waner/RealWorld/Summary3.html , though it may go into much more depth than you need.



.
 
Last edited by a moderator:
  • #3
Other than what K29 has suggested, here are some methods commonly used to solve systems of linear equations (matrix equations). You can do a little research, or refer in your book/s:

1. Cramer's rule.
2. Gauss elimination method (reduce to row echelon form)
3. LU-factorisation method
4. Using matrix inversion.

Personally, i prefer the Gauss elimination method which is quicker.
 
  • #4
K29 said:
Well you could re-arrange it with matrix algebra:

[itex]Ax=b[/itex]

[itex]Axx^{T}=bx^{T}[/itex]

[itex]A=\frac{bx^{T}}{xx^{T}}[/itex]

Note the last step is allowed because [itex]xx^{T}[/itex] is a scalar.(You can't divide matrices by matrices)

Familiarise yourself with the transpose of a matrix (in this case, a column vector) and matrix multiplication, and perhaps rules of matrix algebra and that should be all you need to understand the above.

This link may be helpful http://people.hofstra.edu/stefan_waner/RealWorld/Summary3.html , though it may go into much more depth than you need.



.

This is incorrect. Since x and b are column vectors, the objects xxT and bxT are 3×3 matrices, not scalars.

Anyway, if x and b are known but A is unknown, the equations Ax = b give 3 equations in the 9 unknowns aij, so the system is underdetermined. Additional information or some type of optimization criterion would need to be incorporated in order to obtain a unique solution.

RGV
 
Last edited by a moderator:
  • #5
K29 said:
Well you could re-arrange it with matrix algebra:

[itex]Ax=b[/itex]

[itex]Axx^{T}=bx^{T}[/itex]

[itex]A=\frac{bx^{T}}{xx^{T}}[/itex]

Note the last step is allowed because [itex]xx^{T}[/itex] is a scalar.(You can't divide matrices by matrices)
The last step is disallowed because [itex]xx^{T}[/itex] is a 3x3 matrix. What you might be able to do is post-multiply by the inverse of [itex]xx^{T}[/itex], but that too is disallowed because [itex]xx^{T}[/itex] is singular.

So let's go back to the start.
Genxi said:
I am trying to solve for the A matrix (3 x 3). I know matrix x is (3 x 1) and matrix b is (3 x 1), how do I go about solving for matrix A?
You can't. Ax=b comprises three equations. However, you have nine unknowns, the nine elements of A. That's an underdetermined system. There are either no solutions or there are an infinite number of solutions.Edit
I now see that Ray Vickson beat me to it.
 
  • #6
My mistake. I was working too quickly. Apologies to OP
 

FAQ: Rules for Solving Unknown Matrix A in Ax=b Problem?

What is the meaning of Ax=b if matrix A is unknown?

Ax=b is an equation that represents a system of linear equations, where A is a matrix of coefficients, x is a vector of unknown variables, and b is a vector of constants. The equation represents the relationship between the unknown variables and the constants. If matrix A is unknown, it means that the coefficients in the system of equations are not given, and need to be solved for.

Can we solve for x if matrix A is unknown?

Yes, we can still solve for x even if matrix A is unknown. This process is known as matrix inversion, where the unknown matrix is found by multiplying both sides of the equation by the inverse of matrix A. However, this method may not always be possible if the matrix is singular or ill-conditioned.

How can we find the inverse of matrix A if it is unknown?

In order to find the inverse of matrix A, we need to use a technique called matrix inversion. This involves transforming the original equation Ax=b into the form A^-1Ax=A^-1b, where A^-1 is the inverse of matrix A. By multiplying both sides by A^-1, we can isolate x and solve for it.

Is it possible for matrix A to be unknown in all cases?

No, matrix A cannot be unknown in all cases. If we have the same number of unknown variables and equations in the system, we can find the unique solution for matrix A. However, if the system is underdetermined (more unknown variables than equations) or overdetermined (more equations than unknown variables), then matrix A may be unknown.

Are there any methods to solve for x without knowing matrix A?

Yes, there are other methods to solve for x without knowing matrix A, such as Gaussian elimination or Cramer's rule. These methods involve manipulating the system of equations to solve for the unknown variables without explicitly knowing the matrix of coefficients. However, these methods may not always be efficient or accurate, especially if the system is large and complex.

Back
Top