Gilbert Strang, Professor Emeritus of Mathematics at MIT, taught linear algebra for more than 60 years and is credited with laying the mathematical foundations used by millions of AI engineers. His flagship course, 18.06 Linear Algebra, became one of the most-watched STEM lecture series on MIT OpenCourseWare, and his textbook "Linear Algebra and Learning from Data" explicitly bridges classical matrix methods with deep learning. VnExpress profiles how his course and freely available lectures continue to underpin how data scientists and ML practitioners around the world learn the mathematics of modern AI.
Background
Gilbert Strang joined MIT's mathematics faculty in 1962 and taught linear algebra there for more than 60 years, delivering his final 18.06 lecture in May 2023 to a standing ovation at age 88. He has since held the title of Professor Emeritus.
Why it matters for AI
Linear algebra - covering vector spaces, matrix operations, eigendecomposition, and singular value decomposition - is the core mathematical language of machine learning. Neural network layers are matrix multiplications; backpropagation relies on the chain rule applied to matrix-valued functions; PCA and attention mechanisms in transformers are directly grounded in SVD and inner-product geometry. Strang's courses and textbooks, especially "Linear Algebra and Learning from Data", were among the first to make these connections explicit for a practitioner audience.
Open access reach
MIT OpenCourseWare has made Strang's 18.06 lectures freely available online, where they have been viewed tens of millions of times. That scale of reach has made his curriculum one of the most widely consumed mathematical educational resources in the history of AI and data science.
Editorial note
This is a profile article covering an established educational legacy, not a new product, model, or research result. The immediate practitioner impact is access to the existing free OCW lectures and Strang's books, which remain among the highest-quality introductions to the mathematical foundations of AI.
Scoring Rationale #
Biographical profile of a retired professor whose influence on AI education is well established but historical. No new research, tool, or development is announced. Relevant to practitioners as a reminder of foundational resources, but does not change the state of the field.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.