MIT OCW 18.06: Linear Algebra — Summary & Key Concepts

Instructor: Gilbert Strang Platform: MIT OpenCourseWare Difficulty: Introductory Department: Mathematics Original course: View on MIT OCW

Course Overview

MIT 18.06 is one of the most-watched academic courses in history, and for good reason. Taught by Gilbert Strang — who literally wrote the textbook on linear algebra — this course transforms what many students experience as a dry, formula-heavy subject into a deeply geometric and intuitive discipline. Strang's approach emphasizes the four fundamental subspaces, the geometry of matrix operations, and the central role of eigenvalues in understanding linear systems. Linear algebra is the mathematical backbone of machine learning, computer graphics, quantum mechanics, statistics, and signal processing. Strang's ability to make these connections vivid and accessible is why millions of learners have turned to this course as their definitive introduction to the subject.

Key Concepts

1. Vectors, matrices, and systems of linear equations — The course begins with solving Ax = b through elimination, then reframes systems of equations geometrically. Row operations become matrix factorizations (A = LU), and students see how computational methods connect to structural properties of matrices.

2. The four fundamental subspaces — The column space, null space, row space, and left null space of a matrix form a complete geometric picture of what a linear transformation does. Strang builds the entire course around this framework, showing how every matrix partitions space into these four orthogonal components.

3. Orthogonality and projections — Orthogonal vectors, orthogonal complements, and the Gram-Schmidt process lay the groundwork for projections and least-squares fitting. The projection matrix P = A(A^T A)^{-1} A^T is derived geometrically, giving students the intuition behind regression and signal approximation.

4. Eigenvalues and eigenvectors — An eigenvector is a direction that a matrix stretches or compresses without rotating. Eigenvalues quantify that stretch. This concept unlocks matrix diagonalization, the spectral theorem, and applications ranging from Google's PageRank to quantum mechanics. Strang makes eigenvalues visual and tangible.

5. The Singular Value Decomposition (SVD) — The SVD is the culmination of the course: every matrix, regardless of shape or rank, can be decomposed into orthogonal rotations and a diagonal scaling. This factorization is the most important in all of applied mathematics, powering data compression, dimensionality reduction, and recommender systems.

Module/Lecture Breakdown

Notable Insights

"The column space tells you which right-hand sides b have solutions. The null space tells you what freedom you have in those solutions. Together, they tell you everything about Ax = b." — Gilbert Strang, on the fundamental subspaces

"Eigenvalues are not just numbers that come out of a determinant calculation. They are the heartbeat of a matrix — the rates at which it grows, decays, or oscillates." — Gilbert Strang, on eigenvalues

"The SVD is the climax of linear algebra. Every matrix has one. It reveals the geometry that other factorizations only hint at." — Gilbert Strang, on the Singular Value Decomposition

"Linear algebra has become the most useful subject in mathematics for the 21st century. Data science, machine learning, computer graphics — they all run on matrix computations." — Gilbert Strang, on the relevance of linear algebra

Who Should Take This Course

• First- and second-year university students in math, science, or engineering who need a strong linear algebra foundation
• Machine learning and data science practitioners who want to deeply understand the math behind PCA, SVD, and linear regression
• Self-learners who found linear algebra confusing in school and want a teacher who makes the geometry come alive
• Physics and engineering students who need to understand eigenvalues, orthogonal decompositions, and matrix methods for differential equations
• Anyone who wants to experience one of the greatest mathematics lectures ever recorded, taught by a professor who has dedicated his career to making linear algebra accessible

Build Your Course Knowledge Vault

Taking an online course and want to retain what you learn? DistillNote helps you capture, organize, and review course material — automatically generating structured notes from video lectures and readings.

Paste a lecture URL → get structured summaries, key concepts, and searchable notes in 60 seconds. Build a vault of everything you study.

Try DistillNote free — no credit card required

More MIT OCW course summaries: View all courses Related: AI Course Note Taker · Best Study Tools 2026