STAT 3660

This course trains students to critically apply statistical and computational techniques to real-world data analysis problems, with an emphasis on high-dimensional data. Building on foundations from earlier courses, students will deepen their understanding of how to frame analytical objectives, preprocess complex datasets, and apply appropriate modeling approaches. Topics include exploratory data analysis (i.e., data acquisition, data preprocessing, data cleaning, visualization techniques, numerical summaries, data exploration), dimension reduction methods (linear and nonlinear) for simplifying and interpreting high-dimensional datasets, and supervised and unsupervised learning of both continuous and categorical outcomes. Algorithms may include clustering algorithms, multiple linear regression, logistic regression, decision trees, neural networks and pre-trained models. The course emphasizes not only technical skills but also critical reasoning about model assumptions, generalization, and the implications of statistical learning in applied contexts. PREREQUISITE: CS 1910, STAT 2240 and STAT 2910 Three lecture hours per week