A system of linear equations can be expressed compactly in matrix form as \( AX = B \). This representation organizes the coefficients, variables, and constants of the equations into three distinct matrices.
In this form:
| Symbol | Description |
|---|---|
| \[ A \] | Coefficient Matrix - Contains all coefficients of variables in the system |
| \[ X \] | Variable Vector - Column vector containing all unknown variables |
| \[ B \] | Constant Vector - Column vector containing constants from the right side of equations |
| \[ a_{ij} \] | Coefficient Element - Coefficient of variable j in equation i |
| \[ A^{-1} \] | Inverse Matrix - Exists only when det(A) ≠ 0, used for direct solution |
| \[ [A|B] \] | Augmented Matrix - A and B combined for Gaussian elimination |
| \[ \text{rank}(A) \] | Matrix Rank - Number of linearly independent rows/columns |
The matrix equation \( AX = B \) represents a system of m linear equations with n variables. The components are structured as follows:
The product of the m × n matrix A and the n × 1 vector X results in an m × 1 vector, which must equal the vector B.
A system \( AX = B \) is consistent (has at least one solution) if and only if the rank of the coefficient matrix A is equal to the rank of the augmented matrix [A|B]. This means the constant vector B must be in the column space of A.
For a square system (m=n), a unique solution exists if and only if the determinant of the coefficient matrix is non-zero (det(A) ≠ 0). If det(A) = 0, the system has either no solution or infinitely many solutions.
A system of the form \( AX = 0 \) is called a homogeneous system. It always has the trivial solution \( X = 0 \). Non-trivial solutions exist only if det(A) = 0.
If \( X_1 \) is a solution to \( AX = B_1 \) and \( X_2 \) is a solution to \( AX = B_2 \), then for any scalars c₁ and c₂, the vector \( c_1X_1 + c_2X_2 \) is a solution to \( AX = c_1B_1 + c_2B_2 \). This demonstrates the linearity of the system.
The matrix form \( AX = B \) is derived directly from the definition of matrix multiplication. Consider a simple system of two linear equations with two variables:
We can group the coefficients, variables, and constants into three separate matrices:
Now, let's compute the matrix product AX:
By equating this result with the constant matrix B, we reconstruct the original system of equations. Therefore, the compact form \( AX = B \) is equivalent to the original system.
Engineers use large systems of linear equations in matrix form to model and analyze the forces and stresses in complex structures like bridges, dams, and skyscrapers. This ensures the designs are stable and can withstand various loads.
In circuit analysis, Kirchhoff's laws for current and voltage result in a system of linear equations. Matrix form is used to solve for the unknown currents and voltages in complex circuits with multiple loops and power sources.
Economists model market behavior with systems of equations representing supply, demand, and pricing. Matrix algebra helps find equilibrium points. In finance, it is used for portfolio optimization by solving systems that balance risk and return across multiple assets.
Matrix equations are fundamental to 3D computer graphics for transformations like rotation, scaling, and translation. In robotics, they are used to calculate the positions and movements of robotic arms and joints.
Many machine learning algorithms, such as linear regression, involve solving matrix equations to find the optimal parameters (weights) that best fit a set of data. This is crucial for creating predictive models.
The internet can be modeled as a massive system of equations where each webpage's rank is a variable dependent on the ranks of pages linking to it. Google uses sophisticated matrix methods to solve this enormous system, determining the importance and search order of billions of webpages.
Economists build input-output models to understand how different sectors of an economy (e.g., manufacturing, agriculture, energy) are interrelated. These models form large matrix equations that can predict how a change in one sector, like a drop in oil prices, will ripple through the entire economy.
Your GPS receiver constantly solves systems of linear equations. It receives signals from multiple satellites, each providing an equation relating your unknown position (x, y, z) and time to the satellite's known position. By solving this system in matrix form, the receiver pinpoints your exact location on Earth.
Systems of linear equations represented by \( AX=B \) are classified based on the number of solutions they have. This is primarily determined by the properties of the coefficient matrix A and its relationship with the constant vector B.
| Condition | Solution Type | Description |
|---|---|---|
| `det(A) ≠ 0` (for square matrices) | Unique Solution | The system is consistent and independent. There is exactly one set of values for the variables that satisfies all equations. |
| `det(A) = 0` and `rank(A) = rank([A|B])` | Infinitely Many Solutions | The system is consistent and dependent. At least one equation is a linear combination of the others, leading to an infinite number of possible solutions. |
| `det(A) = 0` and `rank(A) ≠ rank([A|B])` | No Solution | The system is inconsistent. The equations are contradictory, and there is no set of values for the variables that can satisfy all of them simultaneously. |
A special case is the Homogeneous System where \( B = 0 \). These systems always have at least one solution, the trivial solution \( X = 0 \). They have non-trivial (infinite) solutions if and only if `det(A) = 0`.
Incorrect Matrix Setup: A common error is incorrectly transcribing the system into matrix form. Ensure coefficients are in the correct rows and columns in matrix A, variables are in the correct order in vector X, and signs are preserved. A single mistake here will lead to a completely wrong solution.
Assuming an Inverse Always Exists: Students often try to solve for X using \( A^{-1}B \) without first checking if the inverse exists. You must calculate the determinant of A first. If det(A) = 0, the matrix is singular, has no inverse, and this method cannot be used.
Forgetting Matrix Multiplication Order: The solution is \( X = A^{-1}B \), not \( X = BA^{-1} \). Matrix multiplication is not commutative, so the order is critical. Always pre-multiply B by the inverse of A.