- The student will be able to perform a linear regression analysis, understand its limitations and transform the data if necessary to better fit the assumptions
- The student will be able to use graphical tools to judge the outcomes of her analysis and detect outlying and/or influential data points
- The student will be able to fit a linear logistic regression model, and other Generalized Linear Models
- The student is familiar with the theory behind these methods, and is able to use that to test hypotheses and find confidence regions for unknown parameters
- The student is able to apply some basic machine learning techniques for data analysis and is able to use simulation to assess the performance of these models and test hypotheses
|
|
This course is essential for anyone who has to do data analysis involving one or more covariates. The linear model and logistic regression are the most widely used statistical tools, and therefore also the most widely abused tools. When analysing data, a mathematician should be aware of all the pitfalls that could be there. This course intends to make the students aware of this, and offer solutions and alternative methods to analyse data, such as non-parametric shape-restricted regression.
This course has a strong practical component, where students have to do their own analysis. However, there will also be a thorough treatment of the underlying theory.
Instructional Modes
|
|
|
|
Probability and Statistics
|
|
| Written and computer assignment |
|
|