Engineering Math Series 12: Elimination, Inverse Matrices, and Determinants

There are multiple ways to handle systems of linear equations, but at the introductory level it is best to distinguish three perspectives.

1. Elimination as a computational procedure
2. Inverse matrices as a conceptual representation
3. Determinants as an invertibility test

These three are interconnected but serve different roles.

Gaussian Elimination

Gaussian elimination uses row operations to transform the system into a simpler form. It is the most fundamental and practical method in actual computation.

For example, in

$$\begin{aligned} 2x + y &= 5 \\ x - y &= 1 \end{aligned}$$

you can eliminate unknowns one by one by doubling the second equation and subtracting it from the first.

The advantage of elimination is that the computation is systematic and it naturally reveals whether there are no solutions or multiple solutions.

Inverse Matrices

If matrix $A$ is invertible, then

$A^{-1}A = I$

and the solution of

$A\mathbf{x} = \mathbf{b}$

can be written as

$\mathbf{x} = A^{-1}\mathbf{b}$

This expression is very clean, but in actual numerical computation, directly computing the inverse is not always the best approach. Therefore, the inverse matrix is a computational tool, but more importantly it should be understood as a conceptual tool.

Determinants

The determinant is an important number that tests the invertibility of a square matrix. For a 2x2 matrix,

$$\det \begin{pmatrix} a & b \\ c & d \end{pmatrix} = ad - bc$$

If the determinant is not zero, the matrix is invertible; if zero, it is singular.

$$\det(A) \neq 0 \quad \Rightarrow \quad A^{-1} \text{ exists}$$

Worked Example

For the matrix

$$A = \begin{pmatrix} 2 & 1 \\ 1 & -1 \end{pmatrix}$$

the determinant is

$\det(A) = 2(-1) - 1\cdot1 = -3$

Since it is not zero, $A$ is invertible.

The system

$$A\mathbf{x} = \begin{pmatrix} 5 \\ 1 \end{pmatrix}$$

has a unique solution. Solving gives $(x, y) = (2, 1)$.

What matters more than the numbers themselves is that the determinant tells us "does this system converge to a single point."

Engineering Applications

Simulation and Numerical Analysis

Large linear systems are mostly solved by elimination-type algorithms.

Robotics and Graphics

Whether a coordinate transformation matrix is invertible is connected to whether the original state can be recovered.

Data Science

When solving normal equations or linear regression, the conditioning and invertibility of the matrix are very important.

Common Mistakes

Only trying to find the inverse matrix

For small examples it is fine, but in actual computation, elimination or factorization methods are more central.

Memorizing the determinant as "just a formula"

The determinant is not just a calculation problem but information showing how much the transformation distorts space.

Missing the meaning of row operations

Row operations are not changing the equation but moving to an equivalent system with the same solution.

One-Line Summary

Elimination is the solving method, the inverse matrix is a representation tool, and the determinant is an invertibility test -- understanding these roles clarifies the structure.

Next Post Preview

In the next post, we will look at one of the most important ideas in linear algebra: eigenvalues and eigenvectors.

References

• Erwin Kreyszig, Advanced Engineering Mathematics, 10th Edition
• Gilbert Strang, Introduction to Linear Algebra
• Lloyd N. Trefethen, David Bau III, Numerical Linear Algebra