MS-C1620 - Statistical Inference, Lecture, 9.1.2023-19.4.2023
Kurssiasetusten perusteella kurssi on päättynyt 19.04.2023 Etsi kursseja: MS-C1620
Osion kuvaus
-
Welcome to the course Statistical Inference 2023! These MyCourses pages will host the lecture plan, lecture slides, exercise assignments, and other material.
PLEASE NOTE THAT THE REGISTRATION TIME IN SISU ENDS 8.1.2023 23.59!
In the exercise sessions, we will be using R programming language and use R Studio Desktop as the graphical user interface for R.
Note that Introduction to R programming will be lectured as an intensive course on 21.1.-22.1. in case you are interested. See here: https://mycourses.aalto.fi/course/view.php?id=38219#section-0
You can also get assistance with the exercises in the Laskutupa (i.e., mathematics tutoring lab).Lecturer: Jukka Kohonen.
Head assistant: Aleksi Avela. Contact the head assistant (or the assistant of your own group) in questions related to exercise sessions.
When contacting the teachers by e-mail, start by mentioning the course code. (Otherwise how can the teacher know which of their courses you are talking about?) -
The course lectures are on Thursdays at 8.15-10.00 in the hall U2.Exception: The last lecture L12, on 13.4., will actually be groupwork in Zoom (see the table below for links).The lectures will be video recorded, so you can watch them afterwards. Also available are recordings from spring 2022 Zoom lectures. Note that the lecture room videos have 2--3 video streams (camera + one or two screens). You can switch between them when watching from Panopto.On the live lectures, you are encouraged to ask questions. Sometimes the questions may warrant a longer answer than I can provide on the lecture, and I will post extended answers here under the Q&A thread. (The thread also contains questions and answers from the previous year.)In the column "Planned topic", you can find the slides for the planned lecture. Actual lectures may run faster or slower. If the actual lectures are lagging behind, you may want to read the slides on your own, for example, if you need tools for the exercises. Until updated, slides from previous year are available here.Lecture plan
Lecture Day Planned topic, slides Videos etc. Lecture slides
updatedExample code L1 12.1. Descriptive statistics new, old, Q&A 11.1. L1 code L2 19.1. Confidence intervals and hypothesis testing (no new video,
see old) old19.1. L2 code L3 26.1. Non-parametric tests (distribution-free) new, old, Q&A L3 code L4 2.2. Inference for binary data new, old,
extra slidesL4 code L5 9.2. Distribution tests new, old L5 code L6 16.2. Correlation and dependence new, old L6 code Period break L7 2.3. Linear regression I (no new video,
see old), oldL7 code L8 9.3. Linear regression II new (note two screens)
old, Q&AL8 code
+ matlab code showing regression planeL9 16.3. Linear regression III,
Model complexity and Cross validation
lecture cancelled,
read slides and watch old videoold L9 code L10 23.3. ANOVA new, old L10 code L11 30.3. Kernel regression new, old L11 code L12 13.4. Reflection groupwork:
instructions, Zoom link, Miro board, L12 attendanceMiro results
as PDF file
-
1. Exercise groups and assistants. Assignment sheets for the exercises can be found at the bottom of this page.
Group Time Assistant Location Notes H01 Mon 16.15 - 18.00 Alisa Hakola Y344 H02 Tue 10.15 - 12.00 Timo Norrkniivilä Y344 H03 Tue 12.15 - 14.00 Alisa Hakola Y307a Updated for IV-period H04 Tue 14.15 - 16.00 Timo Norrkniivilä U344 H05 Tue 16.15 - 18.00 Aleksi Avela Y344 H06 Wed 10.15 - 12.00 Kristian Jakobsson Y344 H07 Wed 12.15 - 14.00 Mikko Kaivola U6 On 25.1. in U9 H08 Thu 16.15 - 18.00 Mikko Kaivola Maari B (Linux) - 182 H09 Fri 8.15 - 10.00 Jaakko Wallenius U351 H10 Fri 14.15 - 16.00 Jaakko Wallenius Y338 H11 Wed 16.15 - 18.00 Jaakko Wallenius Y344 H12 Wed 8.15 - 10.00 Kristian Jakobsson Y344 H13 Tue 10.15 - 12.00 Jaakko Wallenius Y338 2. Grading
The course grade (0-5) is determined based on the total of exam points (0-24) and exercise points (0-6),
- Total points = exam points + exercise points.
- 15 total points <-> grade 1
- 16 total points <-> grade 2
- 19 total points <-> grade 3
- 22 total points <-> grade 4
- 25 total points <-> grade 5
Additionally, grade 1 is also awarded to those who get 12 points from the exam alone.
Note, especially, that the highest grade (5) can not be obtained without any exercise points. The reason for this is that learning to use R, and learning to conduct statistical analysis in practice are essential learning goals of the course.
The exercise points are valid in all exams organized during the year 2023.If you miss some exercise session (e.g. the first week because of travels), please study the problems on your own. You can still continue the course.Exercises
There are several exercise groups in the course. Please attend one of them weekly and arrive in time in order to get exercise points from the homework.
The homework marks are checked at the beginning of each exercise session with attendance list which is filled at the beginning of each exercise session using MyCourses attendance box. One student is then randomly selected to present their work. The presentation can be R code, powerpoint show, whiteboard drawing, photograph of your hand-written solutions, latex, word, live drawing with paint - which ever you prefer as long as you are confident to give a short explanation of your work (we strongly encourage you to make and present your homework solution with / as R code). Please remember that your solution does not need to be correct - an attempt is enough - and that the assistant is there for you to complement and help to communicate your work if needed You can find the group times and locations in SISU.
The course exercises consist of two kinds of problems, homework problems and class problems.
- Homework problems: the first problem(s) of each week's exercise sheet is a homework problem which must be completed before that week's exercise session. Bring your solution to the homework problems with you to the exercise sessions (exception: first week's exercise). In the beginning of the exercise session, the course assistant will randomly ask some people who have solved the homework problem to present their solutions to everyone.
- Class problems: the remaining problems of each week's exercise sheet are class problems that will be solved together in the exercise sessions.
Additionally, some of the homework and class problems are optional. And although no points will be awarded from them, you are still heavily encouraged to go them through.
Exercise points will be awarded from the sessions as follows:- 1/2 exercise points, if you actively attend the exercise session and had completed the week's homework problem.
- 1/4 exercise points, if you actively attend the exercise session but had not completed the week's homework problem.
If you are unable to attend your own group, you are allowed to attend any of the other groups. If, on a given week, you cannot attend the exercises at all, contact the head assistant (Aleksi Avela) by e-mail. Note that you need an extremely good reason for getting the points without attendance!
3. Assignment sheets
The planned assignment sheets can be found below.
The model solutions will be downloaded in the respective week's folder after the week's last exercise session. -
This page contains the attendance boxes which are used to mark your attendance and homework points in exercise sessions. Please mark your points to the group which you are currently attending even if you have joined into another group in SISU.
- Mark "Present and Homework" if you attend to the exercise session and have done the week's homework
- Mark "Present" if you attend to the exercise session but have not done the week's homework
The homework marks are checked at the beginning of each exercise session and one student is randomly selected to present their work. The presentation can be R code, powerpoint show, whiteboard drawing, photograph of your hand-written solutions, latex, word, live drawing with paint - which ever you prefer as long as you are confident to give a short explanation of your work. Please remember that your solution does not need to be correct - an attempt is enough - and that the assistant is there for you to complement and help to communicate your work if needed.