Matrix Equations

Definition and Notation

Matrix Equation Form

Matrix equation: expression of linear system as A𝑥 = b. A: coefficient matrix (m×n). 𝑥: unknown vector (n×1). b: constant vector (m×1).

Vector and Matrix Notation

Vector 𝑥: column vector. Matrix A: array of elements a_ij with rows i and columns j. b: known vector.

Compact Representation

Matrix equations unify multiple linear equations into single compact multiplication form.

Matrix Multiplication Basics

Definition

Product C = AB defined when columns of A = rows of B. Element c_ij = Σ_k a_ik b_kj.

Properties

Associative: (AB)C = A(BC). Distributive: A(B + C) = AB + AC. Not generally commutative: AB ≠ BA.

Implications for Equations

Order matters: A𝑥 defined, but 𝑥A generally undefined. Proper dimensions critical.

Types of Matrix Equations

Standard Linear Systems

Form: A𝑥 = b. Goal: find 𝑥 satisfying equation.

Matrix Equations with Multiple Unknowns

Form: AX = B, where X and B are matrices. Solve for matrix X.

Generalized and Quadratic Matrix Equations

Examples: AX + XB = C (Sylvester equation), nonlinear forms involving powers or products of X.

Methods for Solving Matrix Equations

Direct Inversion Method

If A invertible, 𝑥 = A⁻¹b. Requires computing inverse matrix.

Gaussian Elimination

Row operations reduce augmented matrix [A|b] to echelon form. Back-substitution retrieves solution.

LU Decomposition

Factor A into L (lower triangular) and U (upper triangular). Solve LY = b, then UX = Y.

Iterative Methods

Jacobi, Gauss-Seidel, Conjugate Gradient for large sparse systems.

Invertible Matrices and Solutions

Definition of Invertibility

Matrix A invertible if ∃ A⁻¹: AA⁻¹ = I = A⁻¹A. I: identity matrix.

Existence and Uniqueness of Solutions

If A invertible, system A𝑥 = b has unique solution 𝑥 = A⁻¹b.

Non-Invertible Cases

If A singular (det(A)=0), solutions may be infinite or none. Requires alternative methods.

Rank, Nullity, and Solution Existence

Matrix Rank

Rank: maximum number of linearly independent rows or columns of A.

Nullity

Nullity: dimension of null space N(A) = {x | A𝑥 = 0}.

Rank-Nullity Theorem

Rank(A) + Nullity(A) = number of columns of A.

Consistency Conditions

System consistent if rank(A) = rank([A|b]). Otherwise, no solution.

Matrix Equations as Linear Transformations

Interpretation

Matrix A represents linear transformation T: V → W; 𝑥 mapped to A𝑥.

Kernel and Image

Kernel (null space): vectors mapped to zero. Image (range): set of all possible outputs.

Relation to Solutions

Solution 𝑥 to A𝑥 = b exists if b in image of A.

Homogeneous vs Non-Homogeneous Systems

Homogeneous Systems

Form: A𝑥 = 0. Always has trivial solution 𝑥 = 0.

Non-Homogeneous Systems

Form: A𝑥 = b, b ≠ 0. Solutions depend on rank and invertibility.

Solution Sets

Homogeneous: solution space is null space of A. Non-homogeneous: solution set = particular solution + null space.

Applications in Systems of Equations

Engineering

Modeling circuits, statics, control systems.

Computer Science

Graphics transformations, machine learning algorithms.

Economics and Statistics

Input-output models, least squares regression.

Physics

Quantum mechanics, system dynamics.

Computational Algorithms

Matrix Factorization Techniques

LU, QR, Cholesky decompositions for efficient solving.

Iterative Solvers

Conjugate gradient, GMRES for large-scale problems.

Numerical Stability

Pivoting, condition number analysis to avoid errors.

Software Libraries

MATLAB, NumPy, LAPACK implement matrix equation solvers.

Illustrative Examples

Example 1: Simple 2x2 System

A = [[2, 1], [5, 3]], b = [1, 2]Solve for x:x = A⁻¹b

Example 2: LU Decomposition

Given A and b, Decompose A = LU,Solve LY = b,Then UX = Y,Find X.

Example 3: Homogeneous System

Find null space of A where A𝑥 = 0.

Common Errors and Misconceptions

Confusing Matrix Multiplication Order

AB ≠ BA in general; order critical.

Assuming All Matrices are Invertible

Singular matrices lack inverses; solution uniqueness lost.

Ignoring Dimensions

Dimension mismatch causes undefined products.

Overlooking Homogeneous Solutions

Null spaces critical for full solution sets.

Miscalculating Rank

Incorrect rank leads to wrong conclusions on solution existence.

References

• Strang, G., Introduction to Linear Algebra, Wellesley-Cambridge Press, Vol. 5, 2016, pp. 1-600.
• Horn, R. A., & Johnson, C. R., Matrix Analysis, Cambridge University Press, Vol. 2, 2012, pp. 1-650.
• Lay, D. C., Linear Algebra and Its Applications, Pearson, Vol. 4, 2015, pp. 1-576.
• Trefethen, L. N., & Bau, D., Numerical Linear Algebra, SIAM, Vol. 1, 1997, pp. 1-400.
• Axler, S., Linear Algebra Done Right, Springer, Vol. 3, 2015, pp. 1-350.