SI 422 - Regression Analysis

Course content
  • Simple and multiple linear regression models – estimation, tests and confidence regions. Simultaneous testing methods- Bonferroni method etc.
  • Analysis of Variance for simple and multiple regression models. Analysis of residuals. Lack of fit tests. Checks (graphical procedures and tests) for model assumptions: Normality, homogeneity of errors, independence, correlation of covariates and errors. Multicollinearity, outliers, leverage and measures of influence.
  • Model selection (stepwise, forward and backward, best subset selection) and model validation.
  • Discussion of algorithms for model selection. Regression models with indicator variables. Polynomial regression models. Regression models with interaction terms. Transformation of response variables and covariates.
  • Variance stabilizing transformations, Box-Cox method. Ridge`s regression. Weighted Regression.
  • Draper, N. and Smith,H. Applied Regression Analysis, 3rd Edition, John Wiley and Sons Series in Probability and Statistics, New York, 1998.
  • Montgomery, D., Peck, E., Vining, G. Introduction to Linear Regression Analysis, 5th Edition, John Wiley, New York, 2013
  • Sen, A. and Srivastava, M. Regression Analysis Theory, Methods & Applications, 1st Edition, Springer-Verlag Berlin Heidelberg, New York, 1990.
  • Kutner, M., Nachtsheim, C., Neter, J. and Li, W. Applied Linear Statistical Models, 5th Edition, McGraw-Hill Companies, Boston, 2005.
Pre-requisite : SI 427 (Exposure) (For students from other departments, instructor’s permission will be required)
Total credits : 8
Type :
Duration : Spring 2023
Name(s) of other Academic units to whom the course may be relevant : N/A