M484 - Regression Analysis

Introduction to simple linear regression, least square estimation and hypothesis testing of model parameters, prediction, interval estimation in simple linear regression, Coefficient of determination, estimation by maximum likelihood, multiple linear regression, matrix representation of the regression model, estimation and testing of model parameters and prediction, model adequacy checking-residual analysis, PRESS statistics, outlier detection, lack of fit test, serial correlation and Durbin-Watson test, transformation and weighting to correct model inadequacies-variance-stabilizing transformation, generalized and weighted least squares, diagnostics for influential observations, Cook’s D test, multicollinearity-sources and effects, diagnosis and treatment for multicollinearity, ridge regression and LASSO, bootstrap estimation, dummy variable model, variable selection and model building–stepwise methods, polynomial regression and interaction regression models, nonlinear regression, generalized linear models-logistic regression and Poisson regression.