Machine Learning Methods in Non-Life Insurance

Course content

Basic theory of penalized linear regression, splines, additive models, neural networks, multivariate adaptive splines, projection pursuit regression, regression trees, random forests, boosting. Also various methods of classification.

 

Education

MSc Programme in Actuarial Mathematics

MSc Programme in Mathematics-Economics

MSc Programme in Statistics

Learning outcome

Knowledge:

  • Standard penalized methods such as ridge regression and the lasso
  • Know splines, additive,  additive models, neural networks, MARS
  • Regression trees, random forest, boosting
  • Classification with classical methods as well as machine learning methods


Skills:

A general ability to use machine learning methods to solve practical problems


Competences:

  • Know how to use R to solve practical problems

4 hours of lectures per week for 7 weeks

Lecture notes

Non-life insurance 2 (Skade 2) or similar. A class in regression is very useful. It is possible to follow the class without these, but of course it will be more demanding.

Academic qualifications equivalent to a BSc degree is recommended.

Collective
ECTS
7,5 ECTS
Type of assessment
Written assignment, 12 timer
12 hour take-home exam. Collaboration not allowed.
Aid
All aids allowed
Marking scale
7-point grading scale
Censorship form
External censorship
One external examiner.
Criteria for exam assessment

The student must in a satisfactory way demonstrate that he/she has mastered the learning outcome of the course.

Single subject courses (day)

  • Category
  • Hours
  • Lectures
  • 28
  • Preparation
  • 124
  • Project work
  • 42
  • Exam
  • 12
  • English
  • 206