Home
Course information:
Instructor: Ana-Maria Staicu [last name pronounced sty:ku]
Office: 5220 SAS Hall
Phone: 919 515-0644
Email: astaicu [at] ncsu [dot] edu
Meeting information: Tu/Th 10:15-11:30AM, 1216 SAS Hall
This course provides an introduction to use of statistical methods for analyzing multivariate data (multiple variables or traits measured for the same individual) and longitudinal data (same variable or trait measured repeatedly on individuals over time) collected in experiments and surveys. Topics covered include multivariate analysis of variance, discriminant analysis, principal components analysis, factor analysis, covariance modeling, and mixed effects models such as growth curves and random coefficient models. Emphasis is on use of a computer to perform statistical analysis of multivariate and longitudinal data. (Previously, students had to take ST 731 and ST 732 in order to cover these topics. This new course will combine topics from those courses and be aimed at students across campus and Statistics Masters students.)
Software used in this course: R is freely available at http://www.r-project.org/. Download and install R. Go to http://cran.r-project.org/ and follow the instructions at the top of the page.
Lectures:
1/9 Introduction Intro
Algebra revision. Matrix_revision. Normal distribution and other common distns. Normal distn [revised, 1/11].
1/16 MLE for Normal distribution. Data display in R.
Multivariate Data Analysis
1/23 Graphics for univariate data. More on graphics for: Univariate/Bivariate/Multivariate
1/25 Inference about mean vector/s. Notes Part I Notes Part II
2/6 Factor methods: Principal Component Analysis Factor Analysis
2/20 Discriminant analysis and Classification
2/29 Midterm I [on applied multivariate analysis]
3/5-3/9 Spring break (no classes)
Longitudinal Data Analysis
3/13 Intro to LDA and General Linear Model
3/27 Linear Mixed Effects (including Random Coefficient Model)
4/10 Review: Generalized Linear Models. Generalized Repeated measures.
4/19 Midterm II [on applied longitudinal data analysis]
4/26 Project presentation (poster)
Grades: HW (here) and Midterm1+Midterm2+Group Project (here)