Abstract

This is a complete 15 ECTS course on Machine Learning. Building on the math knowledge acquired from the course Linear Algebra and Probability, students will be introduced to Machine Learning during the first part of the course. In the second part, recent machine learning research will be addressed.

Description

The course enables students to analyse machine learning algorithms, implement abstractly specified machine learning methods in an imperative programming language, modify machine learning methods to analyse practical datasets and convey the results.

Subjects include probability distributions; linear models for regression; linear models for classification; deep learning; kernel methods; kernel machines; graphical models; clustering; mixture models and expectation maximization; approximate inference; continuous latent variables; sequential data; ensemble methods; and reinforcement learning.

Formal prerequisites

Linear Algebra and Probability

Intended learning outcomes

After the course, the student should be able to:

• Discuss, clearly explain, and reflect upon central machine learning concepts and algorithms.
• Choose among and make use of the most important machine learning approaches in order to apply (match) them to practical problems.
• Implement abstractly specified machine learning methods in an imperative programming language.
• Combine and modify machine learning methods to analyse practical dataset and covey the results.

Learning activities

The first 10 weeks will consist of lectures and exercise sessions. The last 4 weeks will consist of project work where students will implement machine learning methods from a recent research article.

Mandatory activities

The mandatory activities are weekly exercises that the student solve prior to the mandatory exercise session and/or home and will present the solution to the exercise class if randomly picked by the TA. In order to be qualified for the exam, the student must have solved and volunteered to present his/her solution to 50% of the mandatory problems on average. The completion rate is computed from the lists where the student check, prior to the exercise session, those problem he/she has solved and will be ready to present to the class. On the basis of the presented solution, the TA will give feedback and discuss the solution with the class and complement the solution if necessary. The mandatory weekly exercises occur in the first 10 weeks of the course and facilitate continuous learning throughout the course. The second attempt will be provided for the students, who do not pass the mandatories in the first attempt, after the first ten weeks of the course.

The student will receive the grade NA (not approved) at the ordinary exam, if the mandatory activities are not approved and the student will use an exam attempt.

Course literature

Christopher Bishop: Pattern Recognition and Machine Learning, ISBN 978-0-387-31073-2;

and additional material informed on the course page on LearnIT.

Student Activity Budget

Estimated distribution of learning activities for the typical student

• Preparation for lectures and exercises: 6%
• Lectures: 18%
• Exercises: 18%
• Assignments: 18%
• Project work, supervision included: 30%
• Exam with preparation: 10%

Ordinary exam

Exam type:
D: Submission of written work with following oral, External (7-point scale)
Exam variation:
D2G: Submission for groups with following oral exam supplemented by the submission. Shared responsibility for the report.

Time and date

Ordinary Exam - submission Fri, 7 Jan 2022, 08:00 - 14:00
Ordinary Exam Tue, 25 Jan 2022, 09:00 - 21:00
Ordinary Exam Wed, 26 Jan 2022, 09:00 - 21:00
Ordinary Exam Thu, 27 Jan 2022, 09:00 - 21:00
Ordinary Exam Fri, 28 Jan 2022, 09:00 - 21:00