| Simple Linear Model: assumptions, least squares estimation, model diagnostics, transformations, tests,
prediction. Multiple Regression Models: matrix formulation and analysis including estimation and tests of
assumptions, outliers and measures of influence, variable selection. Fitting linear models using
appropriate software and interpretation. Analysis of Variance (ANOVA): One-Way ANOVA, Two-Way
ANOVA.
Limitations of Linear Model. Exponential dispersion family and properties. Derivation of deviance
function, Likelihood ratio tests, AICs. Estimation of parameters: The Fisher iterative algorithm, Hessian
components. Logit and Log Linear Models. Multinomial logit. Applications to real life examples and
interpretation. The special cases of over dispersed and under dispersed models and model comparisons |
Administrative assistant: DILMAHOMED BOCUS Bibi Swaleha
Telephone: 4037400
Email: s.dilmahomed@uom.ac.mu |
