Skip to content
Machine Learning

Linear Algebra for Machine Learning

Account required to view full content

Every machine learning model you will meet in a quant interview, from linear regression to a deep neural network, is built on a small set of linear algebra ideas. Data is stored in matrices, predictions are matrix multiplications, training is an optimization over vectors, and the most important dimensionality-reduction tool in finance (PCA) is an eigenvalue problem. This lesson gives you the working linear algebra you need to reason about models and to answer the linear algebra questions that come up on the whiteboard. We keep the foundational proofs in our Linear Algebra course and focus here on the parts that matter for machine learning and for the interview.