The course introduces methods of statistical inference, useful in any area of science that collects and analyzes data. The course discusses the methodological foundations, as well as issues of practical implementation and use. Methods discussed in this class are applicable to a broad range of problems, from design and analysis of empirical studies of complex real-life phenomena, to design and analysis of evaluations of computer experiments or computer science research. The coursework includes a term project involving programming and/or work with real-life investigations.
The course discusses the following topics:
Basics of frequentist statistical inference for continuous data: measures of association, confidence and prediction intervals, hypothesis testing, benefits and limitations of p-values.
Experimental design: ways to include data to the study to maximize its statistical efficiency, and strategies for analysis of the designed experiments. Factorial and randomized block designs, linear mixed effects models, and response surface exploration.
Statistical inference for categorical data: measures of associations, and generalized linear models.
Introduction to causal inference: graphical models, adjustments for confounders, interventions and counterfactual inference.
Instructor: Prof. Olga Vitek Email: o.vitek@neu.edu Office hours: Tue and Fri after the class, or by appointment. WVH 310F. Mailbox: WVH 202
Teaching assistants: Mr. Sicheng Hao, Email: hao.sic@husky.neu.edu Office hours: Tue 2-3pm, or by appointment. WVH 310.
Course policies and administration: Syllabus, Piazza, Blackboard. Academic integrity policy is strictly enforced.
(KNNL) Applied Linear Statistical Models. Kutner, Neter, Nachtsheim, Li, McGraw-Hill, 5th Edition, 2004. Datasets. Additional texts will be posted dynamically on Piazza