Matrices

Overview

Matrices are the bedrock of modern data science, providing a concise framework for representing data, networks, and transformations. For your CMI Data Science entrance, deep proficiency in matrix algebra—from basic operations to advanced concepts like PCA and SVD—is a critical prerequisite. This chapter equips you with the computational tools to tackle complex datasets and problem-solving scenarios efficiently.

Chapter Contents

| # | Topic | What You'll Learn | |---|---------------------------|---------------------------------------------------| | 1 | Introduction to Matrices | Define matrices and their basic properties. | | 2 | Basic Matrix Operations | Perform addition, subtraction, scalar operations. | | 3 | Matrix Multiplication | Compute products and understand non-commutativity.| | 4 | Transpose of a Matrix | Find transposes and apply their properties. | | 5 | Inverse of a Matrix | Calculate inverses and solve linear systems. |

Part 1: Introduction to Matrices

Matrices represent linear transformations and systems of equations, forming the essential bedrock for advanced topics like eigenvalues, eigenvectors, and SVD.

1. Basic Matrix Types and Notations

:::question type="MSQ" question="Let $U$ be an $n \times n$ upper triangular matrix and $D$ be an $n \times n$ diagonal matrix. Which of the following statements are ALWAYS true?" options=["$U^T$ is a lower triangular matrix.","$U + D$ is an upper triangular matrix.","$U$ must have non-zero elements on its main diagonal.","If $U$ has all zeros on the main diagonal, it is a strictly upper triangular matrix."] answer="A,B,D" hint="Apply the definitions of upper triangular ($u_{ij}=0$ for $i>j$) and diagonal ($d_{ij}=0$ for $i \neq j$) matrices directly to the properties." solution="Option A: $U^T$ swaps the indices $i$ and $j$. Since $u_{ij} = 0$ for $i>j$ in $U$, the new matrix has zeros where $j>i$ (which is above the diagonal), making it lower triangular. (True) Option B: $D$ only has non-zero elements where $i=j$. Adding $D$ to $U$ only affects the main diagonal values. The elements where $i>j$ remain strictly zero. Thus, $U+D$ is upper triangular. (True) Option C: An upper triangular matrix can have zeros anywhere, including on the main diagonal, as long as all elements below the diagonal are zero. (False) Option D: By definition, if all elements below the diagonal AND on the diagonal are zero, it forms a strictly upper triangular matrix. (True) Answer: A, B, D" :::

Part 2: Basic Matrix Operations

Understanding how to manipulate matrices through basic arithmetic is fundamental before progressing to complex transformations.

1. Matrix Addition and Subtraction

Two matrices $A$ and $B$ can be added or subtracted if and only if they have the exact same dimensions ($m \times n$). Core Properties: * Commutativity: $A + B = B + A$ * Associativity: $(A + B) + C = A + (B + C)$ * Additive Identity: $A + 0 = A$ (where $0$ is the zero matrix of the same dimensions) * Closure: The sum of two upper (or lower) triangular matrices is also an upper (or lower) triangular matrix.

2. Scalar Multiplication

Multiplying a matrix $A$ by a scalar $k$ involves multiplying every single element of the matrix by $k$. Core Properties: * Distributivity over Matrix Addition: $k(A + B) = kA + kB$ * Distributivity over Scalar Addition: $(k + l)A = kA + lA$ * Associativity: $k(lA) = (kl)A$ :::question type="NAT" question="Let $A$ and $B$ be $2 \times 2$ matrices where $A = \begin{pmatrix} 2 & -1 \\ 3 & 4 \end{pmatrix}$ and $2A - 3B = \begin{pmatrix} -2 & 4 \\ 0 & 5 \end{pmatrix}$. What is the value of the element $b_{22}$ in matrix $B$?" answer="1" hint="Set up the algebraic equation for the specific element you need. You do not need to compute the entire matrix $B$." solution="Step 1: Isolate the equation for the specific element $b_{22}$. The equation is $2A - 3B = C$. For the element in the 2nd row, 2nd column: $2a_{22} - 3b_{22} = c_{22}$. Step 2: Substitute the known values. $a_{22} = 4$ $c_{22} = 5$ $2(4) - 3b_{22} = 5$ Step 3: Solve for $b_{22}$. $8 - 3b_{22} = 5$ $-3b_{22} = -3$ $b_{22} = 1$ Answer: 1" :::

Part 3: Matrix Multiplication

Matrix multiplication is the cornerstone of linear algebra operations. It represents the composition of linear transformations and is structurally essential for algorithms ranging from neural network forward passes to solving complex systems of equations, graph theory analysis, and data projections.

1. Core Properties of Multiplication

2. Special Vector Products

Vectors are specialized matrices. A column vector $x$ in $\mathbb{R}^n$ is an $n \times 1$ matrix, and its transpose $x^T$ is a $1 \times n$ row vector. * Inner Product (Dot Product): $x^T y = \sum_{i=1}^{n} x_i y_i$. This results in a $1 \times 1$ scalar. It geometrically represents projection and alignment. * Outer Product: $x y^T$. This results in an $n \times n$ rank-1 matrix where the $(i,j)$-th element is $x_i y_j$. This is heavily used in covariance matrix calculations and SVD.

3. Matrix Exponentiation ($A^n$)

Multiplying a square matrix by itself $n$ times ($A^n$) is a frequent operation in analyzing Markov chains, system dynamics, and recursion. To compute high powers, look for these structural shortcuts: * Diagonal Matrices: If $D$ is a diagonal matrix, exponentiation is applied element-wise. If $D = \text{diag}(d_1, d_2, \dots, d_k)$, then $D^n = \text{diag}(d_1^n, d_2^n, \dots, d_k^n)$. * Rotation Matrices: A 2D counter-clockwise rotation matrix by angle $\theta$ is $R(\theta) = \begin{pmatrix} \cos\theta & -\sin\theta \\ \sin\theta & \cos\theta \end{pmatrix}$. Applying it $n$ times yields a rotation by $n\theta$: $R(\theta)^n = R(n\theta) = \begin{pmatrix} \cos(n\theta) & -\sin(n\theta) \\ \sin(n\theta) & \cos(n\theta) \end{pmatrix}$. * Nilpotent Matrices: A matrix $N$ is nilpotent if $N^k = 0$ for some integer $k$. For a matrix $A$ that can be decomposed as $A = I + N$, we can use the binomial expansion (since $I$ and $N$ commute): $(I+N)^n = I + nN + \frac{n(n-1)}{2}N^2 + \dots$. This series naturally terminates at $k-1$.

4. Computational Complexity (FLOPs)

Understanding the hardware cost of matrix operations is a staple in Data Science. A floating-point operation (FLOP) includes additions and multiplications.

5. Graph Theory & Adjacency Matrices

Matrix multiplication is the mathematical engine behind graph theory and network analysis. For a graph with $m$ vertices, an adjacency matrix $A$ is an $m \times m$ matrix where $A_{ij} = 1$ if there is a directed edge from vertex $i$ to vertex $j$, and $0$ otherwise. Key Analytical Property: The element $(i,j)$ of the matrix $A^k$, denoted as $A^k(i,j)$, represents the exact number of distinct paths of length $k$ from vertex $i$ to vertex $j$.

6. The Trace Property

The trace of a square matrix is the sum of its main diagonal elements: $\text{tr}(A) = \sum a_{ii}$. The trace interacts uniquely with multiplication. Cyclic Permutation: For any matrices $A$ (size $m \times n$) and $B$ (size $n \times m$), both $AB$ and $BA$ result in square matrices (though of different sizes), and: $\text{tr}(AB) = \text{tr}(BA)$ Note: This extends strictly to cyclic permutations for multiple matrices. For example, $\text{tr}(ABC) = \text{tr}(BCA) = \text{tr}(CAB)$. However, it does NOT generally equal non-cyclic permutations like $\text{tr}(ACB)$.

7. Quadratic Forms

A quadratic form maps a vector to a scalar using a symmetric matrix, which is foundational in optimization and defining cost functions. For a symmetric matrix $A$ and a vector $x$, the quadratic form is: $Q(x) = x^T A x$ If $A$ is positive definite, $x^T A x > 0$ for all non-zero vectors $x$. --- :::question type="MSQ" question="Let $A$ and $B$ be two $n \times n$ non-zero matrices. Which of the following statements are ALWAYS true?" options=["If $AB=0$, then both $A$ and $B$ must be singular (non-invertible) matrices.","$\text{tr}(A^T B) = \text{tr}(AB^T)$","$(A+B)^2 = A^2 + 2AB + B^2$","If $A$ is an adjacency matrix of an unweighted graph, $A^3(i,i)$ gives the exact number of distinct triangles passing through vertex $i$ (ignoring direction)."] answer="A,B" hint="Recall the zero product property of matrices and trace cyclic properties. Be careful with matrix algebra expansion, as commutativity cannot be assumed." solution="Option A: Suppose $A$ is invertible. We can multiply the equation $AB=0$ by $A^{-1}$ on the left to get $B=0$. However, the premise states $B$ is a non-zero matrix, creating a contradiction. Therefore, $A$ cannot be invertible. By symmetric logic, $B$ cannot be invertible. Both must be singular. (True) Option B: Using the property $\text{tr}(X) = \text{tr}(X^T)$ and the transpose product rule $(XY)^T = Y^T X^T$, we can evaluate the first expression: $\text{tr}(A^T B) = \text{tr}((A^T B)^T) = \text{tr}(B^T (A^T)^T) = \text{tr}(B^T A)$. Finally, using the cyclic property of trace $\text{tr}(XY) = \text{tr}(YX)$, we know $\text{tr}(B^T A) = \text{tr}(AB^T)$. (True) Option C: Expansion gives $(A+B)(A+B) = A^2 + AB + BA + B^2$. Since $AB \neq BA$ in general matrix algebra, this does not neatly simplify to $A^2 + 2AB + B^2$. (False) Option D: $A^3(i,i)$ gives the total number of paths of length 3 from vertex $i$ back to vertex $i$. While a triangle ($i \to j \to k \to i$) is a path of length 3, $A^3(i,i)$ also counts degenerate paths (like $i \to j \to i \to i$ if self-loops exist) and, for undirected graphs, it traverses loops in both directions. It requires further scaling and division (usually by 2 or 6) to yield the exact number of distinct geometric triangles. (False) Answer: A, B" :::

Part 4: Transpose of a Matrix

The transpose operation reorients a matrix by interchanging its rows and columns. It is the foundational operation for defining symmetry, analyzing quadratic forms, and simplifying complex matrix equations.

1. Core Properties of the Transpose

Mastering these properties is non-negotiable, as they are frequently used to manipulate and simplify abstract matrix expressions in CMI questions.

2. Symmetric Matrices

A square matrix $A$ is symmetric if $A^T = A$ (meaning $a_{ij} = a_{ji}$). Key Properties: * If $A$ and $B$ are symmetric, $A+B$ and $kA$ are symmetric. * The product $AB$ of two symmetric matrices is symmetric if and only if they commute ($AB = BA$). For any* matrix $X$ (not necessarily square), the matrices $X X^T$ and $X^T X$ are ALWAYS symmetric. For any* square matrix $X$, the matrix $X + X^T$ is ALWAYS symmetric.

3. Antisymmetric (Skew-Symmetric) Matrices

A square matrix $A$ is antisymmetric (or skew-symmetric) if $A^T = -A$ (meaning $a_{ij} = -a_{ji}$). Key Properties: * Zero Diagonal: All main diagonal entries of an antisymmetric matrix must be exactly zero (since $a_{ii} = -a_{ii} \implies 2a_{ii} = 0$). * Squares: If $A$ is antisymmetric, its square $A^2$ is symmetric. Sum/Difference: For any* square matrix $X$, the matrix $X - X^T$ is ALWAYS antisymmetric. * Determinant (CRITICAL): If $A$ is an $n \times n$ antisymmetric matrix and $n$ is odd, then $|A| = 0$ (it is strictly non-invertible). If $n$ is even, its determinant is a perfect square.

4. Transpose of Block Matrices

When transposing a block matrix, you must transpose each individual block AND swap the positions of the off-diagonal blocks. If $M = \begin{pmatrix} A & B \\ C & D \end{pmatrix}$, then $M^T = \begin{pmatrix} A^T & C^T \\ B^T & D^T \end{pmatrix}$. --- :::question type="MCQ" question="Let $A$ be an $m \times n$ matrix and $B$ be an $n \times p$ matrix. Which of the following statements is ALWAYS true?" options=["$(A^T B^T)^T = AB$","$(A^T B)^T = B^T A$","$(BA)^T = A^T B^T$","$(AB)^T = B^T A^T$"] answer="$(AB)^T = B^T A^T$" hint="Recall the transpose property for matrix products. Pay attention to the order of multiplication." solution="The property $(AB)^T = B^T A^T$ is the fundamental reversal rule for the transpose of a product. Option A expands to $BA$, which $\neq AB$. Option B expands to $B^T A$, which is only valid if dimensions allow it, but doesn't relate to the standard rule. Option C has the wrong order. Answer: $(AB)^T = B^T A^T$." ::: :::question type="MSQ" question="Let $A$ and $B$ be $n \times n$ matrices. Which of the following statements are ALWAYS true?" options=["If $A$ and $B$ are symmetric, then $AB$ is always symmetric.","If $A$ is antisymmetric, then $A^2$ is symmetric.","If $A$ is an antisymmetric $5 \times 5$ matrix, then $A$ is invertible.","For any square matrix $X$, the matrix $X+X^T$ is symmetric."] answer="B,D" hint="Apply transpose properties to each expression. Remember the odd-dimension determinant rule for skew-symmetric matrices." solution="Option A: $(AB)^T = B^T A^T = BA$. For $AB$ to be symmetric, $AB$ must equal $BA$, which is not generally true. (False) Option B: $(A^2)^T = (AA)^T = A^T A^T = (-A)(-A) = A^2$. Thus, $A^2$ is symmetric. (True) Option C: For an antisymmetric matrix $A$ of odd dimension $n=5$, $|A| = |-A| = (-1)^5|A| = -|A| \implies 2|A| = 0 \implies |A| = 0$. It is non-invertible. (False) Option D: Let $S = X+X^T$. $S^T = (X+X^T)^T = X^T + (X^T)^T = X^T + X = S$. (True) Answer: B, D" ::: :::question type="MSQ" question="Let $P$ and $Q$ be two $n \times n$ skew-symmetric matrices. Which of the following matrices are GUARANTEED to be skew-symmetric?" options=["$P + Q$","$PQ$","$PQ - QP$","$P^3$"] answer="A,C,D" hint="Use the definition $P^T = -P$ and $Q^T = -Q$. Apply the product transpose rule carefully." solution="Option A: $(P+Q)^T = P^T + Q^T = -P - Q = -(P+Q)$. Skew-symmetric. (True) Option B: $(PQ)^T = Q^T P^T = (-Q)(-P) = QP$. This is symmetric only if $QP=PQ$, but not skew-symmetric. (False) Option C: $(PQ - QP)^T = (PQ)^T - (QP)^T = Q^T P^T - P^T Q^T = (-Q)(-P) - (-P)(-Q) = QP - PQ = -(PQ - QP)$. Skew-symmetric. (True) Option D: $(P^3)^T = (P P P)^T = P^T P^T P^T = (-P)(-P)(-P) = -P^3$. Skew-symmetric. (True) Answer: A, C, D" :::

Part 5: Inverse of a Matrix

Introduction

The inverse of a matrix is a fundamental concept in linear algebra, analogous to the reciprocal of a non-zero scalar. It plays a crucial role in solving systems of linear equations, understanding linear transformations, and various matrix decompositions. For a square matrix, its inverse, if it exists, allows us to "undo" the transformation represented by the original matrix. In the context of CMI, understanding matrix inverses is essential for analyzing the solvability of linear systems, properties of matrix transformations, and for advanced topics such as eigenvalues and eigenvectors. This section will thoroughly cover the definition, existence conditions, methods of computation, and key properties of matrix inverses, preparing you to tackle complex problems efficiently. ---

Key Concepts

1. Existence and Uniqueness of the Inverse

For a matrix inverse to exist, the matrix must first be square. Not all square matrices have an inverse. A critical condition for the existence of $A^{-1}$ is that the determinant of $A$ must be non-zero. If an inverse exists, it is unique. This can be proven as follows: Proof of Uniqueness: Assume that a square matrix $A$ has two inverses, say $B$ and $C$. Step 1: By definition of an inverse, we have: $AB = BA = I$ $AC = CA = I$ Step 2: Consider the product $BAC$. We can group this product in two ways. $(BA)C = IC = C$ $B(AC) = BI = B$ Step 3: Equating the two results from Step 2, we find: $C = B$ This proves that if an inverse exists, it must be unique. ---

2. Inverse of a $2 \times 2$ Matrix

For a $2 \times 2$ matrix, there is a straightforward formula to calculate its inverse. This formula is frequently tested, especially in questions involving properties of matrices. Worked Example: Problem: Find the inverse of the matrix $A = \begin{bmatrix} 3 & 1 \\ 5 & 2 \end{bmatrix}$. Solution: Step 1: Calculate the determinant of $A$. $|A| = (3)(2) - (1)(5) = 6 - 5 = 1$ Since $|A| = 1 \neq 0$, the inverse exists. Step 2: Apply the $2 \times 2$ inverse formula. $A^{-1} = \frac{1}{1} \begin{bmatrix} 2 & -1 \\ -5 & 3 \end{bmatrix}$ Step 3: Simplify the expression. $A^{-1} = \begin{bmatrix} 2 & -1 \\ -5 & 3 \end{bmatrix}$ Answer: $A^{-1} = \begin{bmatrix} 2 & -1 \\ -5 & 3 \end{bmatrix}$ ---

3. General Method for Inverse: Adjugate Formula

For matrices of order $n > 2$, the adjugate method (also known as the adjoint method) is a general formula for finding the inverse. This method involves calculating the determinant, minors, cofactors, and the adjugate matrix. Worked Example: Problem: Find the inverse of the matrix $A = \begin{pmatrix} 1 & 2 & 0 \\ 0 & 1 & 1 \\ 0 & 0 & 1 \end{pmatrix}$. Solution: Step 1: Calculate the determinant of $A$. For an upper triangular matrix, the determinant is the product of its diagonal elements. $|A| = (1)(1)(1) = 1$ Since $|A| = 1 \neq 0$, the inverse exists. Step 2: Calculate the cofactors $C_{ij}$ for each element $a_{ij}$. $C_{11} = (-1)^{1+1} \det \begin{pmatrix} 1 & 1 \\ 0 & 1 \end{pmatrix} = 1(1-0) = 1$ $C_{12} = (-1)^{1+2} \det \begin{pmatrix} 0 & 1 \\ 0 & 1 \end{pmatrix} = -1(0-0) = 0$ $C_{13} = (-1)^{1+3} \det \begin{pmatrix} 0 & 1 \\ 0 & 0 \end{pmatrix} = 1(0-0) = 0$ $C_{21} = (-1)^{2+1} \det \begin{pmatrix} 2 & 0 \\ 0 & 1 \end{pmatrix} = -1(2-0) = -2$ $C_{22} = (-1)^{2+2} \det \begin{pmatrix} 1 & 0 \\ 0 & 1 \end{pmatrix} = 1(1-0) = 1$ $C_{23} = (-1)^{2+3} \det \begin{pmatrix} 1 & 2 \\ 0 & 0 \end{pmatrix} = -1(0-0) = 0$ $C_{31} = (-1)^{3+1} \det \begin{pmatrix} 2 & 0 \\ 1 & 1 \end{pmatrix} = 1(2-0) = 2$ $C_{32} = (-1)^{3+2} \det \begin{pmatrix} 1 & 0 \\ 0 & 1 \end{pmatrix} = -1(1-0) = -1$ $C_{33} = (-1)^{3+3} \det \begin{pmatrix} 1 & 2 \\ 0 & 1 \end{pmatrix} = 1(1-0) = 1$ Step 3: Form the cofactor matrix $C$. $C = \begin{pmatrix} 1 & 0 & 0 \\ -2 & 1 & 0 \\ 2 & -1 & 1 \end{pmatrix}$ Step 4: Find the adjugate matrix $\text{adj}(A)$ by transposing the cofactor matrix. $\text{adj}(A) = C^T = \begin{pmatrix} 1 & -2 & 2 \\ 0 & 1 & -1 \\ 0 & 0 & 1 \end{pmatrix}$ Step 5: Apply the adjugate formula for the inverse. $A^{-1} = \frac{1}{|A|} \text{adj}(A) = \frac{1}{1} \begin{pmatrix} 1 & -2 & 2 \\ 0 & 1 & -1 \\ 0 & 0 & 1 \end{pmatrix}$ Answer: $A^{-1} = \begin{pmatrix} 1 & -2 & 2 \\ 0 & 1 & -1 \\ 0 & 0 & 1 \end{pmatrix}$ ---

4. Properties of Inverse Matrices

Understanding the properties of inverse matrices is crucial for simplifying expressions and solving matrix equations efficiently. Explanation of Key Properties: * Inverse of a Product: This property is particularly important. The order of multiplication is reversed when taking the inverse of a product. * Proof: We need to show that $(B^{-1}A^{-1})(AB) = I$ and $(AB)(B^{-1}A^{-1}) = I$. * $(B^{-1}A^{-1})(AB) = B^{-1}(A^{-1}A)B = B^{-1}IB = B^{-1}B = I$ * $(AB)(B^{-1}A^{-1}) = A(BB^{-1})A^{-1} = AIA^{-1} = AA^{-1} = I$ * Inverse of a Transpose: This property links the operations of transposing and inverting. * Proof: We know $AA^{-1} = I$. Transposing both sides: $(AA^{-1})^T = I^T \implies (A^{-1})^T A^T = I$. This shows that $(A^{-1})^T$ is the inverse of $A^T$. * Determinant of an Inverse: This is a direct consequence of the determinant product rule. * Proof: We have $AA^{-1} = I$. Taking the determinant of both sides: $|AA^{-1}| = |I|$. * Using the property $|AB| = |A||B|$, we get $|A||A^{-1}| = 1$. * Therefore, $|A^{-1}| = \frac{1}{|A|}$. ---

5. Invertibility and Systems of Linear Equations

The existence of a matrix inverse is directly related to the solvability and uniqueness of solutions for systems of linear equations. Consider a system of $n$ linear equations in $n$ variables, represented in matrix form as: $Ax = b$ where $A$ is an $n \times n$ matrix, $x$ is an $n \times 1$ column vector of variables, and $b$ is an $n \times 1$ column vector of constants. Explanation: If $A^{-1}$ exists, we can multiply both sides of $Ax = b$ by $A^{-1}$ from the left: $A^{-1}(Ax) = A^{-1}b$ $(A^{-1}A)x = A^{-1}b$ $Ix = A^{-1}b$ $x = A^{-1}b$ This demonstrates that if $A^{-1}$ exists, the solution $x$ is uniquely determined. This connection is fundamental in numerical methods for solving linear systems. ---

6. Similarity Transformations

A similarity transformation is a transformation of a matrix $B$ into a matrix $A$ such that $A = PBP^{-1}$ for some invertible matrix $P$. This concept is crucial for understanding matrix diagonalization and canonical forms. Properties under Similarity Transformation: Similar matrices share many important properties. * Determinant: Similar matrices have the same determinant. * Proof: $|A| = |PBP^{-1}| = |P||B||P^{-1}| = |P||B|\frac{1}{|P|} = |B|$. * Invertibility: If $A$ is similar to $B$, then $A$ is invertible if and only if $B$ is invertible. * Proof: If $B$ is invertible, then $B^{-1}$ exists. Then $A = PBP^{-1}$ implies that $A^{-1} = (PBP^{-1})^{-1} = (P^{-1})^{-1}B^{-1}P^{-1} = PB^{-1}P^{-1}$. Thus, $A$ is invertible. * Conversely, if $A$ is invertible, then $B = P^{-1}AP$. Following the same logic, $B^{-1} = (P^{-1}AP)^{-1} = P^{-1}A^{-1}(P^{-1})^{-1} = P^{-1}A^{-1}P$. Thus, $B$ is invertible. * Rank: Similar matrices have the same rank. * Trace: Similar matrices have the same trace. * Eigenvalues: Similar matrices have the same eigenvalues. ---

Problem-Solving Strategies

---

Common Mistakes

---

Practice Questions

:::question type="MCQ" question="Let $A = \begin{pmatrix} 2 & 5 \\ 1 & 3 \end{pmatrix}$. Which of the following statements about $A^{-1}$ is true?" options=["$A^{-1} = \begin{pmatrix} 3 & -5 \\ -1 & 2 \end{pmatrix}$","$A^{-1} = \begin{pmatrix} -2 & -5 \\ -1 & -3 \end{pmatrix}$","$A^{-1} = \begin{pmatrix} 2 & -1 \\ -5 & 3 \end{pmatrix}$","The inverse does not exist."] answer="$A^{-1} = \begin{pmatrix} 3 & -5 \\ -1 & 2 \end{pmatrix}$" hint="Use the formula for the inverse of a $2 \times 2$ matrix." solution="Step 1: Calculate the determinant of $A$. $|A| = (2)(3) - (5)(1) = 6 - 5 = 1$ Step 2: Apply the $2 \times 2$ inverse formula $A^{-1} = \frac{1}{|A|} \begin{pmatrix} d & -b \\ -c & a \end{pmatrix}$. $A^{-1} = \frac{1}{1} \begin{pmatrix} 3 & -5 \\ -1 & 2 \end{pmatrix} = \begin{pmatrix} 3 & -5 \\ -1 & 2 \end{pmatrix}$ The correct option is $A^{-1} = \begin{pmatrix} 3 & -5 \\ -1 & 2 \end{pmatrix}$." ::: :::question type="NAT" question="If $A$ is an $n \times n$ invertible matrix and $|A|=4$, what is the value of $|(2A^T)^{-1}|$?" answer="0.03125" hint="Use properties of determinants and inverses: $|kA|=k^n|A|$ and $|A^{-1}|=1/|A|$." solution="Step 1: Use the property $|kA| = k^n|A|$ for a scalar $k$ and $n \times n$ matrix $A$. $|2A^T| = 2^n |A^T|$ Step 2: Use the property $|A^T| = |A|$. $|2A^T| = 2^n |A|$ Step 3: Substitute the given value $|A|=4$. $|2A^T| = 2^n \cdot 4$ Step 4: Use the property $|B^{-1}| = \frac{1}{|B|}$. $|(2A^T)^{-1}| = \frac{1}{|2A^T|} = \frac{1}{2^n \cdot 4}$ Assuming a standard $3 \times 3$ matrix context for this problem, $n=3$. $|(2A^T)^{-1}| = \frac{1}{2^3 \cdot 4} = \frac{1}{8 \cdot 4} = \frac{1}{32}$ As a decimal: $\frac{1}{32} = 0.03125$" ::: :::question type="MSQ" question="Let $A$ and $B$ be $n \times n$ invertible matrices. Which of the following statements is/are necessarily true?" options=["$(A^{-1}B)^T = B^T (A^T)^{-1}$","$(A+B)^{-1} = A^{-1} + B^{-1}$","$A(BA)^{-1}B = I_n$","If $A^2=I$, then $A^{-1}=A$"] answer="$A(BA)^{-1}B = I_n$,If $A^2=I$, then $A^{-1}=A$" hint="Carefully apply the properties of inverse and transpose, paying attention to the order of operations. Consider counterexamples for false statements." solution="1. $(A^{-1}B)^T = B^T (A^T)^{-1}$: Using the property $(XY)^T = Y^T X^T$, we get $(A^{-1}B)^T = B^T (A^{-1})^T$. Using $(A^{-1})^T = (A^T)^{-1}$, we get $B^T (A^T)^{-1}$. (True)

$(A+B)^{-1} = A^{-1} + B^{-1}$: Matrix inversion is not distributive over addition. Counterexample: $A = I, B = I$. Then $(A+B)^{-1} = (2I)^{-1} = 0.5I$, but $A^{-1} + B^{-1} = I + I = 2I$. (False)

$A(BA)^{-1}B = I_n$: Using $(XY)^{-1} = Y^{-1}X^{-1}$, we get $A(A^{-1}B^{-1})B = (AA^{-1})(B^{-1}B) = I_n I_n = I_n$. (True)

If $A^2=I$, then $A^{-1}=A$: If $A^2=I$, then $A \cdot A = I$. By definition of an inverse, $A^{-1}=A$. (True)

The correct options are $A(BA)^{-1}B = I_n$ and If $A^2=I$, then $A^{-1}=A$." ::: :::question type="SUB" question="Given an invertible matrix $A$, prove that $(A^k)^{-1} = (A^{-1})^k$ for any positive integer $k$." answer="Proof by induction" hint="Use mathematical induction. Establish the base case for $k=1$. Assume it holds for $k=m$, then prove it for $k=m+1$ using the property $(AB)^{-1} = B^{-1}A^{-1}$." solution="We will prove this by mathematical induction on $k$. Base Case ($k=1$): For $k=1$, the statement is $(A^1)^{-1} = (A^{-1})^1$. This simplifies to $A^{-1} = A^{-1}$, which is trivially true. Inductive Hypothesis: Assume that the statement holds for some positive integer $m$, i.e., $(A^m)^{-1} = (A^{-1})^m$. Inductive Step ($k=m+1$): We need to show that $(A^{m+1})^{-1} = (A^{-1})^{m+1}$. Step 1: Express $A^{m+1}$ as a product. $A^{m+1} = A^m \cdot A$ Step 2: Apply the property $(XY)^{-1} = Y^{-1}X^{-1}$. $(A^{m+1})^{-1} = (A^m \cdot A)^{-1} = A^{-1} (A^m)^{-1}$ Step 3: Apply the inductive hypothesis. $A^{-1} (A^m)^{-1} = A^{-1} (A^{-1})^m$ Step 4: Combine the terms using exponent rules. $A^{-1} (A^{-1})^m = (A^{-1})^{m+1}$ By the principle of mathematical induction, the statement $(A^k)^{-1} = (A^{-1})^k$ is true for all positive integers $k$." ::: :::question type="MCQ" question="Let $A$ be a $3 \times 3$ matrix such that $A^3 = I$, where $I$ is the $3 \times 3$ identity matrix. Which of the following is equal to $A^{-1}$?" options=["$A$","$A^2$","$I$","$A^4$"] answer="$A^2$" hint="Multiply $A^3=I$ by $A^{-1}$ from the left or right, or recognize the definition of inverse." solution="Given the equation $A^3 = I$. We can rewrite $A^3$ as $A^2 \cdot A = I$. By the definition of an inverse, if $BA = I$, then $B = A^{-1}$. In this case, $B = A^2$. Therefore, $A^{-1} = A^2$. Alternatively, multiply both sides by $A^{-1}$ from the right: $A^3 A^{-1} = I A^{-1} \implies A^2 = A^{-1}$. The correct option is $A^2$." ::: ---

Summary

---

What's Next?

---

Chapter Summary

---

Chapter Review Questions

:::question type="MCQ" question="Let $A = \begin{pmatrix} 1 & 2 \\ 0 & 1 \end{pmatrix}$ and $B = \begin{pmatrix} 1 & 0 \\ 1 & 1 \end{pmatrix}$. Which of the following is equal to $(AB^T)^{-1}$?" options=["A) $\begin{pmatrix} 1 & -3 \\ 0 & 1 \end{pmatrix}$","B) $\begin{pmatrix} 1 & 3 \\ 0 & 1 \end{pmatrix}$","C) $\begin{pmatrix} 1 & -2 \\ 0 & 1 \end{pmatrix}$","D) $\begin{pmatrix} 1 & 2 \\ 0 & 1 \end{pmatrix}$"] answer="A) $\begin{pmatrix} 1 & -3 \\ 0 & 1 \end{pmatrix}$" hint="First, find $B^T$. Then calculate the product $AB^T$. Finally, find the inverse of the resulting matrix." solution="First, find the transpose of matrix $B$: $B^T = \begin{pmatrix} 1 & 1 \\ 0 & 1 \end{pmatrix}$ Next, calculate the product $AB^T$: $AB^T = \begin{pmatrix} 1 & 2 \\ 0 & 1 \end{pmatrix} \begin{pmatrix} 1 & 1 \\ 0 & 1 \end{pmatrix} = \begin{pmatrix} (1)(1)+(2)(0) & (1)(1)+(2)(1) \\ (0)(1)+(1)(0) & (0)(1)+(1)(1) \end{pmatrix} = \begin{pmatrix} 1 & 3 \\ 0 & 1 \end{pmatrix}$ Let $C = AB^T = \begin{pmatrix} 1 & 3 \\ 0 & 1 \end{pmatrix}$. To find $C^{-1}$, we use the formula for a $2 \times 2$ matrix inverse: For $C$, $a=1, b=3, c=0, d=1$. Determinant is $ad-bc = 1-0 = 1$. $C^{-1} = \frac{1}{1} \begin{pmatrix} 1 & -3 \\ 0 & 1 \end{pmatrix} = \begin{pmatrix} 1 & -3 \\ 0 & 1 \end{pmatrix}$ Thus, $(AB^T)^{-1} = \begin{pmatrix} 1 & -3 \\ 0 & 1 \end{pmatrix}$." ::: :::question type="NAT" question="Let $A = \begin{pmatrix} 2 & x \\ 3 & 4 \end{pmatrix}$ and $B = \begin{pmatrix} 1 & 5 \\ 2 & 1 \end{pmatrix}$. If the matrix $A+B$ is symmetric, find the value of $x$." answer="0" hint="A matrix $M$ is symmetric if $M = M^T$. First, find the sum $A+B$, then apply the condition for symmetry." solution="First, calculate the sum $A+B$: $A+B = \begin{pmatrix} 2 & x \\ 3 & 4 \end{pmatrix} + \begin{pmatrix} 1 & 5 \\ 2 & 1 \end{pmatrix} = \begin{pmatrix} 3 & x+5 \\ 5 & 5 \end{pmatrix}$ For a matrix $M$ to be symmetric, its transpose $M^T$ must be equal to $M$. Let $M = A+B = \begin{pmatrix} 3 & x+5 \\ 5 & 5 \end{pmatrix}$. Then its transpose is $M^T = \begin{pmatrix} 3 & 5 \\ x+5 & 5 \end{pmatrix}$. For $M$ to be symmetric, $M = M^T$: $\begin{pmatrix} 3 & x+5 \\ 5 & 5 \end{pmatrix} = \begin{pmatrix} 3 & 5 \\ x+5 & 5 \end{pmatrix}$ Equating the corresponding elements, we get: $x+5 = 5 \implies x = 0$" ::: :::question type="MCQ" question="Given $A = \begin{pmatrix} 3 & 2 \\ 1 & 1 \end{pmatrix}$ and $B = \begin{pmatrix} 5 \\ 2 \end{pmatrix}$. If $AX=B$, what is $X$?" options=["A) $\begin{pmatrix} 1 \\ 1 \end{pmatrix}$","B) $\begin{pmatrix} 1 \\ 2 \end{pmatrix}$","C) $\begin{pmatrix} 2 \\ 1 \end{pmatrix}$","D) $\begin{pmatrix} -1 \\ 2 \end{pmatrix}$"] answer="A) $\begin{pmatrix} 1 \\ 1 \end{pmatrix}$" hint="To solve for $X$ in $AX=B$, you need to pre-multiply both sides by $A^{-1}$." solution="We are given the matrix equation $AX=B$. To find $X$, we need to calculate $A^{-1}$ and then compute $X = A^{-1}B$. First, find the inverse of $A = \begin{pmatrix} 3 & 2 \\ 1 & 1 \end{pmatrix}$. The determinant of $A$ is $\det(A) = (3)(1) - (2)(1) = 3 - 2 = 1$. $A^{-1} = \frac{1}{1} \begin{pmatrix} 1 & -2 \\ -1 & 3 \end{pmatrix} = \begin{pmatrix} 1 & -2 \\ -1 & 3 \end{pmatrix}$ Now, substitute $A^{-1}$ and $B$ into $X = A^{-1}B$: $X = \begin{pmatrix} 1 & -2 \\ -1 & 3 \end{pmatrix} \begin{pmatrix} 5 \\ 2 \end{pmatrix} = \begin{pmatrix} (1)(5) + (-2)(2) \\ (-1)(5) + (3)(2) \end{pmatrix} = \begin{pmatrix} 5 - 4 \\ -5 + 6 \end{pmatrix} = \begin{pmatrix} 1 \\ 1 \end{pmatrix}$ The correct answer is A." ::: :::question type="NAT" question="Let $P = \begin{pmatrix} 1 & 2 \\ 3 & 4 \end{pmatrix}$ and $Q = \begin{pmatrix} 0 & 1 \\ 2 & 3 \end{pmatrix}$. If $2P - Q^T = R$, find the value of $R_{12}$ (the element in the first row, second column of $R$). " answer="2" hint="First, calculate $2P$. Then find the transpose of $Q$, $Q^T$. Finally, perform the subtraction $2P - Q^T$ to find $R$ and identify the element $R_{12}$." solution="First, calculate $2P$: $2P = 2 \begin{pmatrix} 1 & 2 \\ 3 & 4 \end{pmatrix} = \begin{pmatrix} 2 & 4 \\ 6 & 8 \end{pmatrix}$ Next, find the transpose of $Q$: $Q^T = \begin{pmatrix} 0 & 2 \\ 1 & 3 \end{pmatrix}$ Now, calculate $R = 2P - Q^T$: $R = \begin{pmatrix} 2 & 4 \\ 6 & 8 \end{pmatrix} - \begin{pmatrix} 0 & 2 \\ 1 & 3 \end{pmatrix} = \begin{pmatrix} 2-0 & 4-2 \\ 6-1 & 8-3 \end{pmatrix} = \begin{pmatrix} 2 & 2 \\ 5 & 5 \end{pmatrix}$ The element $R_{12}$ is the element in the first row, second column of $R$. From the matrix $R = \begin{pmatrix} 2 & 2 \\ 5 & 5 \end{pmatrix}$, we see that $R_{12} = 2$." ::: ---

What's Next?