IAM574 MATHEMATICAL FOUNDATIONS FOR DATA ANALYSIS
Course Content
Background in linear algebra: vector space, orthogonality, positive definiteness, eigenvalues and eigenvectors; Matrix decomposition: Cholesky decomposition, OR decomposition, singular-value decomposition; Least square problems: projectors; Best low rank approximation: Eckart-Young Theorem, principal component analysis; Vector Calculus: gradients, high order derivatives, chain rule; Continuous optimization: convexity, gradient descent, Lagrange multipliers, stochastic gradient descent; Basics of probability and distribution: Maximum likelihood estimation.