Modern statistics and simulation in Action

Target audience and purpose

The course is suitable for employees in the private sector enjoying working with data and visualization techniques. Whether you already have practical experience with data handling and statistical methods but want to get a more solid theoretical foundation and gain insight into the latest trends, or if you are new to the field and feel that you need a solid theoretical and practical foundation to stand on, the course fits to you.

Implementation

The training is carried out at low pace during the fall semester, with lectures via zoom. You will be supported in implementing your new knowledge through personal coaching on three occasions. You may work on your own data and problems. You may use whatever programming language you want, otherwise we will be using Python and R.

The course is adapted to your needs, ensuring value for you and your business. The lectures are given together with students and are in English. Both group and individual assignments will be used. The coaching will take place on individual basis.

Course content

Module 1: Statistical data analysis

Theory
Probability, conditional probability, Bayes' theorem, discrete and continuous random variables, probability function, distribution function, density function, averages, dispersion measures, multidimensional random variables, dependence measures.

Practice
Data processing with programming or statistical software, data reduction, sparsity and compression, principal component analysis, cluster analysis, machine learning.

Module 2: Statistical inference

Theory
Random sample, sample distributions (t and F distributions), methods for parameter estimation (least squares method, maximum likelihood method), calculation of point and interval estimates for relevant parameters, variance analysis (ANOVA) and variance reduction.

Practice
Inverse transform sampling, implementation of parameter estimates with controlled variance, comparison of estimates based om maximum likelihood method (or other methods for parameter estimation) and estimates based on machine learning.

Course structure

The course will be divided into two parts:

• lectures, where we cover theoretical ideas and concepts
• practical implementations and vizualizations, where we focus on problem solving.

References

G. Casella, R. Berger, (2002). Statistical inference (2nd). Duxbury Press
S. L. Brunton, J.L. Nathan Kutz (2022). Data-driven Science and Engineering: Machine Learning, Dynamical Systems, and Control (2nd). Cambridge University Press

Course date

Course is conducted from September 5 to October 24, including subsequent individual coaching on three one-hour occasions.

Lecturers

Adrian Muntean, prof. dr. habil., has a strong background in mathematical modeling and simulation of real-world processes, specialized in industrial mathematics.

• Researcher profile: https://www.kau.se/en/researchers/adrian-muntean

Grigor Nika, dr., is a researcher in applied analysis and a dedicated teacher passioned about statistics.

• Researcher profile: https://www.kau.se/en/researchers/grigor-nika

Nicklas Jävergård, MSc,  is a physicist by training with strong background in scientific computing. Nicklas is a PhD student in applied mathematics and likes playing with data.

• Researcher profile: https://www.kau.se/personal/nicklas-javergard

Course fee

The course fee is SEK 9.800 excl. VAT per person. Compulsory literature are not included in the fee.

VAT: Swedish VAT, 25%, will be charged on the course fee for all foreign or Swedish participants except for Swedish governmental organisations.

We will send an invoice shortly after the course has started.

Terms of registration 

• Commissioned education is conducted in accordance with its regulations and turn to employees in the public sector, authorities and companies. 
• To register for a commissioned education you have to be employed at a company or organization (not sole proprietorship). 
• The registration is binding.  
• You can not participate as a private person. 
• When you register your employer has to approve to your participation and will agree to receive an invoice for the costs of the education. 
• If the number of participants becomes too low we might have to cancel the education. In that case we let you know as soon as possible. We need five participants and maximum ten participants.

TREFFEKT® – SKILLS DEVELOPMENT WITH EFFECT!

Knowledge only has value if it is used. Our brand Treffekt® means that we are three parties; Karlstad University, you as a participant and your manager/organization who together create the conditions for the knowledge from the education to be used and make an impact in your business. Treffekt® is based on the factors that research shows are the most important for "transfer of learning", i.e. how knowledge from an education is put into practice.

• August 29, 13:15-14:00, Treffekt®-dialog on zoom