Type of course

Admission requirements

A relevant undergraduate Bachelor degree in Engineering program in computer science or equivalent.

In addition, the following requirements must be met:

- minimum 25 credits in mathematics (equivalent to Mathematical Methods 1, 2 og 3), 5 credits in statistics and 7,5 ects i physics on a higher level is required.

Courses at the Master degree program in Computer Science.

Recommended Prerequisite: Basic course in mathematical statistics.

Course content

• The students will have an introduction in time series and regressions supporting predictions based on historic data.
• Some techniques related to image analysis and language models will be introduced and pertinent applications explored
• Techniques like imputations, dimensionality reductions and embeddings will be introduced.
• The course will explore in depth a selection of neural networks such as LSTM, CNN. GAN and Autoencoders.
• The concept of learning agents in a complex, infinite and continuous action and state space will be discussed and explored.
• The students will be introduced to advanced methods for large scale data capture with emphasis on data lakehouse with semantic data catalogues relevant for machine learning

Recommended prerequisites

Objectives of the course

After passing this course the student should have obtained the following learning outcome:

Knowledge

The student shall have;

• Deep insight in a selection of regression, clustering and vectorisation tasks for use with machine learning
• Acquaintance with advanced neural network methods applied to time series, images and text
• Knowledge of advanced agent-based learning, multi-agent systems (MAS)
• Understanding of MAS-methods on a practical problem
• Hands-on knowledge in using a data scientific methods on practical problems
• Understanding of data collection methods for machine learning

Skills

The student should be able to;

• To set up data scientific/machine learning projects
• To include and use advanced software libraries for machine learning and MAS in their own program
• To interpret machine learning results with respect to application and business impact
• To specify a full-fledged architecture for practical, machine learning oriented data collection and management

General Competence

• The student will be posed for industrial artificial intelligence and machine learning projects
• Be able to to build apt agent learning systems
• Be capable of creating smaller distributed control and optimization systems with MAS
• Be able to enter industrial projects on Big Data

Language of instruction and examination

Teaching methods

Up to 30 hours introductory lectures.

Guided exercises

Individual project run by the students under guidance by a supervisor.

Teaching and examination Language:

Lectures will be held in English (Provided the presence of English speaking students, English will be chosen).

Lecture notes will be given in English.

Project reports must be written in the English language.

Information to incoming exchange students

This course is open for inbound exchange student who meets the admission requirements. Please see the Admission requirements" section".

Master Level

Do you have questions about this module? Please check the following website to contact the course coordinator for exchange students at the faculty: https://en.uit.no/education/art?p_document_id=510412.