MATH 662: Mathematics of Machine Learning and Industrial Applications I

Basic mathematical and probabilistic models and derivations for linear and logistic regression including regularization and application to SVM and PCA. Mathematical and numerical aspects of classical learning methods such as Kernel methods andgradient based methods including neural networks. Incorporates modern tools such as Python, shell tools, and version control. Includes industrial scale applications in satellite imagery, physics, biology and engineering. Computational and analytic assignments are given. Offered by Mathematics. Limited to three attempts.