Alan Edelman and Steven G. Johnson
Matrix Calculus For Machine Learning — seventeen lectures from MIT, IAP 2023.Matrix Calculus For Machine Learning And Beyond. Seventeen lectures from MIT’s IAP 2023 course.
Matrix Calculus For Machine Learning · 17 lectures
Part 1: Introduction and Motivation
Part 2: Derivatives as Linear Operators
Part 1: Derivatives in Higher Dimensions
Part 2: Vectorization of Matrix Functions
Part 1: Kronecker Products and Jacobians
Part 2: Finite-Difference Approximations
Part 1: Gradients and Inner Products in Other Vector Spaces
Part 2: Nonlinear Root Finding, Optimization
Part 1: Derivative of Matrix Determinant and Inverse
Part 2: Forward Automatic Differentiation via Dual Numbers
Part 3: Differentiation on Computational Graphs
Part 1: Adjoint Differentiation of ODE Solutions
Part 2: Calculus of Variations and Gradients of Functionals
Part 1: Derivatives of Random Functions
Part 2: Second Derivatives, Bilinear Forms, and Hessian Matrices
Part 1: Derivatives of Eigenproblems
Part 2: Automatic Differentiation on Computational Graphs