STAT 447

Based on a foundation of mathematical and statistical theory, the course covers a series of statistical methods for supervised learning and unsupervised learning, focusing on applications to real data using statistical software. The topics include: resampling methods such as Cross-Validation and Bootstrap; regression and classification including Linear Regression, Logistic Regression, Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), K-Nearest Neighbors (KNN); model selection and regularization including Best Set Selection, Lasso, Elastic Net; non-linear models including Generalized Additive Models (GAM); tree-based methods including Decision Trees, Bagging, Random Forest; Support Vector Machines (SVM); dimension reduction and clustering including Principle Component Analysis (PCA), K-Means, Hierarchical Clustering; Ensemble Learning including Boosting, Stacking, Multi-Task Prediction; and an introduction to Deep Learning.