# About MATH3714

This module is **MATH3714 Linear Regression and Robustness**. The
module manager and lecturer is Dr Jochen Voss, and my email address is
J.Voss@leeds.ac.uk.

## Notes and videos

The main way I expect you to learn the material for this course is by reading these notes and by watching the accompanying videos. I will release two sections of notes each week, for a total of 22 sections.

Reading mathematics is a slow process. Each section roughly corresponds to one traditional lecture, which would have taken 50 minutes. If you find yourself regularly getting through sections in much less than an hour, you’re probably not reading carefully enough through each sentence of explanation and each line of mathematics, including understanding the motivation as well as checking the accuracy.

It is possible (but not recommended) to learn the material by only reading the notes and not watching the videos. It is not possible to learn the material by only watching the videos and not reading the notes.

Since we will all be relying heavily on these notes, I’m even more keen than usual to hear about errors mathematical, typographical or otherwise. Please, please email me if think you may have found any.

## Lectures

There will be one online synchronous “lecture” session each week, on Mondays at 2-3pm, with me, run through Microsoft Teams. These will not be “lectures” in the traditional sense of the term, but will be an opportunity to re-emphasise material you have already learned from notes and videos, to give extra examples, and to answer common student questions, with some degree of interactivity.

I will assume you have completed all the work for the previous week by the time of the lecture, but I will not assume you’ve started the work for that week itself.

I am very keen to hear about things you’d like to go through in the lectures; please email me with your suggestions.

## Workshops and Problem Sheets

There will be 5 problem sheets, corresponding to workshops in weeks 2, 4, 6, 8 and 10. The main goal of the workshops will be to go over your answers to the problems sheets.

My recommended approach to problem sheets and workshops is the following:

- Work through the problem sheet before the workshop, spending plenty
of time on it, and making multiple efforts at questions you get
stuck on. I recommend spending
*at least three hours*on each problem sheet, in more than one block. Collaboration is encouraged when working through the problems, but I recommend writing up your work on your own. - Take advantage of the workshops to ask for help or clarification on questions you weren’t able to complete.
- After the workshop, attempt again the questions you were previously stuck on.
- If you’re still unable to complete a question after this second
round of attempts,
*then*consult the solutions.

## Discussion Board

I have set up a Microsoft Team for the course. I propose to use the “Discussion” channel there as a discussion board. This is a good place to post questions about material from the course, and — even better! — to help answer your colleagues’ questions. The idea is that you all as a group should help each other out. I will visit a couple of times a week to clarify if everybody is stumped by a question, or if there is disagreement.

## Software

For the module we will use the statistical computing package R. This program is free software, and you can find the program and documentation at the R project homepage. In particular, R will be used in the (assessed) practial.

My recommendation would be to install the RStudio environment, which includes R, on your own computer and use this for your work. (Choose the open source version, “RStudio Desktop”, on the download page.) Alternatively you can use RStudio or plain R on the university computers.

## Assessments

Your final mark for the module will be based on a computer practical (20%) and a final exam (80%). For the practical (I believe it will take place in week 10) you will need to solve some problem using R and the methods you learned in the course and to present your results in a short report.