OPTIMIZATION FOUNDATIONS OF DATA SCIENCE

Credit Hours: 3 · Graduate

Optimization methods for machine learning. Topics included are as follows: basics of optimization theory, gradient-based optimization (e.g., gradient descent, stochastic gradient descents, AdaGrad, Adam, RMSProp, etc.), linear regression and its extensions (e.g., ridge regression and lasso), least-squares classification and logistic regression, Newton methods in machine learning, basics of constrained optimization, Lagrangian relaxation and duality, support vector machines, and optimization in neural networks.