Courses Detail Information
STAT4130J – Applied Regression Analysis using R
Instructors:
Jing Liu;Ms. Jiang;Ms. Weng;Ms. Wang;Ms. Wang;Ailin Zhang
Credits:
4 credits
Pre-requisites:
ECE4010J Obtained Credit
Description:
Regression modeling is one of the most useful statistical techniques available. It helps to establish the relationship between a variable of interest (the response variable) and one or more other variables thought to influence the response variable (explanatory variables). Discovering and analyzing such relationships is the basis for understanding the natural world. Consequently, regression modeling is useful in science, medicine, business, finance, marketing and countless other fields.
This course introduces students to modeling the relationship between a response variable and several explanatory variables via linear regression models. Topics include simple and multiple linear regressions; the least squares algorithm for estimation of parameters; hypothesis testing and prediction; model diagnostics and improvement; algorithms for variable selection; nonlinear regression and other methods. Students will use R to automate statistical analyses in this course
Course Topics:
1. Simple and multiple linear regression
2. Geometry and distribution theory of least squares
3. ANOVA, confidence and prediction intervals
4. Model diagnostics and assumption checking
5. Multicollinearity and influential observations
6. Model selection and regularization (AIC, BIC, Lasso, Ridge)
7. Generalized linear models (logistic, Poisson, etc.)
8. Nonlinear and time-series regression
9. Computational algorithms and matrix methods in R