This course aims to explore the art of statistical analysis through models. During the course, you will learn how to explore data and estimate simple linear regression, multivariate, and logistic models. We will discuss these models' finite and asymptotic properties, as well as variable selection and shrinkage methods. All of this will be addressed through a mix of theory and application in R. Statistical modelling is not just about mathematics or elegant models; one must be curious and ask many questions. Even the simplest case can become complex if data is not explored before proceeding.
Learning Outcomes
On completion of the module students should be able to:
- calculate confidence intervals for parameters and prediction intervals for future observations;
- understand how to represent a linear model in matrix form;
- check model assumptions and identify influential observations;
- deal with violations of model assumptions (e.g. non-normal errors, heteroschedasticity, structural breaks in the data);
- understand how outliers and leverage points can affect regression;
- analyse and interpret the parameters and regression output in R;
- have a basic understanding of the Design of Experiments (ANOVA one-way layout);
- understand how to use logistic regression.
- Module Supervisor: Danilo Petti