### Conduct of the course

This is a self-study version of the course. The course will be lectured again in autumn 2020 in period I.

The course will be evaluated based on computer exercises returned to Moodle (50% of the grade) and a home exam (50% of the grade).

You need to obtain at least 50% of the computer exercise points (60/120) in order to participate at the home exam.

For more detailed instructions, please see the course Moodle area. In this self-study version, you can work on the problems at your own pace until 14 August. There will be a home exam at the end of the course between 14 August and 28 August.

### Description

The course is compulsory for students of the Statistics study track in the Master's Programme in Mathematics and Statistics.

Master's Programme in Mathematics and Statistics is responsible for the course.

The course belongs to the Statistics and Social statistics module in the Master's Programme in Mathematics and Statistics.

The course also belongs to the Machine Learning and Statistical Data Science modules in the Master's Programme in Data Science.

The course is available to students from other degree programmes.

BSc courses on linear algebra, probability calculus, statistical

inference; basic programming skills; Bayesian inference (for example

MAT22005 Bayes-päättely, DATA11006 Statistical Data Science or

LSI35002 Bayesian inference in biosciences or similar)

Fundamentals of differential equations

Knowledge and use of common general computational tools to perform reliable statistical analyses. Understanding the theoretical foundations of the most important Monte Carlo methods. Applying and implementing computational statistical procedures on a high-level programming language.

Recommended time/stage of studies for completion: 1. or 2. year of MSc studies.

Term/teaching period when the course will be offered: yearly during the autumn term, period I.

Most important numerical and computational methods and principles for statistics. Theory and practice of methods for sampling from probability distributions including rejection sampling, importance sampling, generic Markov chain Monte Carlo and Hamiltonian Monte Carlo. Overview of modern methods for approximate inference. The computer projects can be implemented in Python (preferred) or R.

Exercises and home exam.

Lecture notes and articles to be announced during the course

Lectures, exercises, computer exercises, project work

Computer excercises and a computer-based home exam, Course will be graded with grades 1-5

Exercises and project work.