MIT OpenCourseWare · Gilbert Strang · 35 video lectures and reviews
Gilbert Strang’s course on linear algebra: solving systems of equations, the subspaces that come with a matrix, projections and least squares, determinants, eigenvalues, positive definite matrices and the singular value decomposition. It suits anyone who needs the mathematics under data science, machine learning, engineering or physics.
One session a lecture: park a question at the minute it came up, answer it from the book or the problem set, and it becomes a recall card. The reviews before each quiz have sessions of their own.
Lecture 1 · Seeing linear equations as rows and as columns
Lecture 2 · Elimination, step by step with matrices
Lecture 3 · Ways to multiply matrices, and finding inverses
Lecture 4 · Writing elimination as A = LU
Lecture 5 · Transposes, reordering rows, and what a vector space is
Lecture 6 · Two key subspaces: the b I can reach, and the x sent to zero
Lecture 7 · Solving Ax = 0: pivots and the unknowns left open
Lecture 8 · Solving Ax = b: the complete solution
Lecture 9 · When vectors are independent, and counting dimensions
Lecture 10 · The four subspaces every matrix comes with
Lecture 11 · Spaces of matrices, rank one, and graphs
Lecture 12 · Graphs and networks as matrices
Lecture 13 · Review before the first quiz
Lecture 14 · Perpendicular vectors and subspaces
Lecture 15 · Projecting onto a subspace
Lecture 16 · Projection matrices, and fitting data by least squares
Lecture 17 · Making columns orthonormal with Gram-Schmidt
Lecture 18 · What a determinant must do
Lecture 19 · Computing determinants: the big formula and cofactors
Lecture 20 · Determinants at work: inverses, Cramer’s rule and volume
Lecture 21 · Eigenvectors: directions a matrix only stretches
Lecture 22 · Diagonalizing a matrix, and taking its powers
Lecture 23 · Differential equations and the matrix exponential
Lecture 24 · Matrices for probabilities (Markov), and Fourier series
Lecture 24b · Review before the second quiz
Lecture 25 · Symmetric matrices, and when all the pivots are positive
Lecture 26 · Complex matrices and the fast Fourier transform
Lecture 27 · Positive definite matrices and finding a minimum
Lecture 28 · When two matrices are similar, and the Jordan form
Lecture 29 · Breaking any matrix into rotations and stretches (the SVD)
Lecture 30 · Linear transformations and the matrices behind them
Lecture 31 · Changing basis, and compressing images
Lecture 32 · Review before the third quiz
Lecture 33 · One-sided inverses and the pseudoinverse
Lecture 34 · Looking back over the whole course
18.06 is from MIT OpenCourseWare, shared under CC BY-NC-SA 4.0. This logbook is Soundings’ own: session names and questions written for it, nothing copied from the course.