Applications of Linear Algebra

The chapters so far have dealt with the mathematics itself; this one turns to look at where it is used. But the striking thing is not how different the applications are — it is how similar they are.

Machine learning treats data as a matrix and a network layer as an affine transformation; computer graphics treats a vertex as a vector and a scene graph as a composition of matrices; finite element analysis in structural engineering reduces a bridge to the solution of Ku=fK\mathbf{u} = \mathbf{f}; and quantum mechanics identifies an observable with a matrix whose eigenvalues are the possible measurement outcomes. Four completely different disciplines, and yet the vocabulary is identical.

The reason is that all of them are asking the same kind of question. “What does this transformation do to this direction?” is a question about eigenvectors. “What is the closest vector in this subspace?” is a question about projection. “How does this system balance out?” is a question about solving a system of linear equations. Once you have learned to recognise these questions, you will find linear algebra turning up in places that seem very far from a textbook.

Chapters

Found an issue with this content?

Tell us what could be improved.