This course introduces common methods and algorithms used in machine learning. Lectures focus on
supervised learning with an emphasis on using current cross-validation methods. Supervised topics
include a variety of linear regression methods, including ordinary, subset, and shrinkage. Supervised
linear models are revisited in the context of classification and extended to quadratic discriminate
analysis. Basis expansions and kernel smoothers are also explored in the regression and classification
settings, along with classification/regression trees, and support vector machines. Finally, if time
permits, unsupervised methods include cluster analysis, principal components, and independent
component analysis. In all instances, the methods will be applied to data sets with widely varying
topics. To succeed in this course, students need decent knowledge of statistics (such as probability
theory, intermediate statistics, or regression analysis) and programming skills (intermediate level of R
or Python) prior to enrolling. All students are encouraged to bring their laptops to every class.