Optimization in Data Science

Course content

In data science, we can split many problems into two parts. The first part concentrates on finding a class of models that fits well to a data generating process in the real world. In the second part, we then fit the model to the data, which often involves some optimization. The topic of this course is optimization. We derive theory on optimization problems and learn about efficient methodology. We learn how to recognize whether an optimization problem is easy or hard and how to transform problems to have a standard form. Optimization problems arise frequently in many different fields but applications in data science will be our main motivation.

Education

MSc Programme in Statistics

Learning outcome

 Knowledge:

  • convex sets and functions
  • duality
  • generalized inequalities
  • optimization algorithms
  • subgradients


Skills:

  • recognizing convex sets and functions
  • applying convex relaxations
  • solving linear and quadratic programs 
  • using optimization software


Competences:

  • recognizing and transforming optimization problems
  • solving different types of optimization problems 
  • relating optimization to statistics

4 hours lectures and 4 hours of exercises per week for 7 weeks.

Basic knowledge of probability theory and regression, e.g. MI, MatStat or equivalent courses. Basic knowledge of programming in R.

Academic qualifications equivalent to a BSc degree is recommended.

ECTS
7,5 ECTS
Type of assessment
Oral examination, 25 minutes
There will be a 30 min preparation time before the oral exam.
Aid
All aids allowed
Marking scale
7-point grading scale
Censorship form
No external censorship
Several internal examiners.
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
  • Theory exercises
  • 28
  • Preparation
  • 115
  • Exam
  • 35
  • English
  • 206