Maths I β Week 8: Linear Algebra
870 words
4 min read
View
π Learning OS Architecture & Ecosystem
The Grand Blueprint: The Learning OS is a rigid, consistent, single-source-of-truth ecosystem designed to eliminate context switching and focus purely on mastery. Every course strictly adheres to this contract.
Maths I β Week 8: Linear Algebra
Short description. Linear Algebra is the mathematics of organized data. Matrices allow us to solve complex systems of equations efficiently by treating entire sets of numbers as single algebraic units.
1. Core Concept
Definition: A Matrix is a rectangular array of numbers arranged in m rows and n columns. Matrix algebra provides a deterministic framework for transformations and solving systems of linear equations.
Intuition: Imagine a matrix as a spreadsheet that can be multiplied by other spreadsheets. It represents a "transformation" of spaceβrotating, scaling, or flipping vectors.
Formula / Rule:
AmΓnββ
BnΓpβ=CmΓpβ
where the element cijβ is the dot product of the i-th row of A and the j-th column of B.
Important
Commutativity Fail: Unlike scalar multiplication, ABξ =BA in most cases. The order of multiplication is structurally sacred.
2. Pattern A β Matrix Multiplication
What to recognize: A request to find the product of two arrays A and B.
Abstract Solution (Strategy)
- [Dimension Audit]: Verify if the number of columns in A exactly matches the number of rows in B.
- [Dot Product Rule]: Each entry cijβ is the sum of products of corresponding elements from Row i of A and Column j of B.
- [Watch for]: Inadvertently swapping the order or incorrectly adding instead of multiplying.
Procedure
- Step 1: Check dimensions (mΓn and nΓp). The outer numbers (m,p) define the result size.
- Step 2: For each coordinate (i,j) in the result, multiply Row i of A with Column j of B element-wise and sum them.
- Step 3: Repeat until the result matrix is saturated.
Worked Example:
Question: Find AB for A=[13β24β] and B=[57β68β].
- Step 1: 2Γ2 times 2Γ2 = 2Γ2 result.
- Step 2: Top-Left coordinate (1,1)=(1β 5)+(2β 7)=5+14=19.
- Step 3: Fill other coordinates similarly.
- Answer: [1943β2250β]
3. Pattern B β 2Γ2 Matrix Inversion
What to recognize: A question asking for Aβ1 where A is a square 2Γ2 matrix.
Abstract Solution (Strategy)
- [Determinant Check]: Calculate Delta=adβbc. If Delta=0, the inverse does not exist (Matrix is singular).
- [Adjugate Swap]: Swap the main diagonal elements (aβd) and negate the off-diagonal elements (βb,βc).
- [Scalar Scaling]: Multiply the resulting adjugate matrix by Ξ1β.
Procedure
- Step 1: Compute D=(m11ββ m22β)β(m12ββ m21β).
- Step 2: Rearrange: [acβbdβ]β[dβcββbaβ].
- Step 3: Multiply by 1/D and simplify.
Worked Example:
Question: Inverse of [25β13β].
- Step 1: det=(2β 3)β(1β 5)=1.
- Step 2: Swap/Negate β[3β5ββ12β].
- Step 3: 11ββ [3β5ββ12β].
- Answer: [3β5ββ12β]
4. Common Mistakes
| Mistake | Why it happens | Correct approach |
|---|---|---|
| Multiplying matrices of incompatible sizes. | Ignoring the dimension audit. | Always check if Col Count of A = Row Count of B. |
| Forgetting to divide by the determinant in inversions. | Skipping the final scalar scaling step. | The inverse is ONLY the adjugate IF the determinant is exactly 1. |
| Assuming AB=BA. | Applying scalar habits to matrix algebra. | The order is fundamental. (AB)T=BTAT. |
5. Flashcards
<Flashcard front="What is the condition for a matrix to be 'singular'?" back="A matrix is singular if its determinant is exactly zero, meaning it cannot be inverted." /> <Flashcard front="How do you compute the Transpose of a matrix?" back="Swap rows for columns: the element at (i, j) moves to (j, i)." /> <Flashcard front="State the Identity Matrix property for multiplication." back="AI = IA = A for any square matrix A." />6. Practice Targets
- Attempt Maths I Graded Assignment 8.
- Calculate the determinant of a 3Γ3 matrix using cofactor expansion on the 1st row.
- Solve Ax=b for 2Γ2 using x=Aβ1b.
7. Connections
| Connects to | How |
|---|---|
| Week 2 β Polynomials | Solving systems of linear equations is the foundational step for partial fraction decomposition and curve fitting. |
| Data Science Core | Matrices are the primary data structure for all machine learning models (tensors). |