MATH 465 Mathematical Foundations of Data Science

Review of eigenvalues and eigenvectors. Fundamental subspaces, matrix factorization techniques, principal components and best lowrank matrices; the structure of neural nets for deep learning. Convergence concepts and limit theorems in probability, basic inequalities of probability, tail bounds, the concentration of measures phenomena, empirical process. Maximum likelihood estimation, regularized regression, the Lasso and its variations. Optimization methods, gradient descent, stochastic gradient descent, convolutional neural nets.
Credit units: 3 ECTS Credit units: 5, Prerequisite:
(MATH 230 or MATH 250 or MATH 255) and (MATH 220 or MATH 223 or MATH 225 or MATH 241).



Bilkent University Main Page
Last regenerated automatically on October 11, 2024 by OAC  Online Academic Catalog Software

