Courses I currently teach or have developed at the University of Iceland. Roughly 300+ students per academic year across these courses.
MSc · English
Master's introduction to deep learning: from the basics of feed-forward networks and back-propagation through modern architectures, with a focus on practical implementation in PyTorch. Includes coverage of convolutional networks, recurrent and attention models, and the basics of training large models. Taught in English.
MSc · Icelandic
Master's course co-taught at the School of Humanities on the use of large language models for Icelandic. Covers what current models can and cannot do for an under-resourced language, evaluation and bias considerations, and practical work with Icelandic-language models and datasets developed at the University of Iceland. Taught in Icelandic.
MSc · English
NLP module taught as part of the Business Intelligence master's programme at the School of Business. Covers the practical NLP toolkit modern analysts need: text preprocessing, embeddings, classification, named-entity recognition, and the use of large language models for business applications. Taught in English.
MSc · English
Master's course developed and taught by Hafsteinn on the full lifecycle of an applied AI project: framing the problem, sourcing and curating data, model selection and evaluation, deployment, and post-deployment monitoring. Students work on team projects that take an idea through every stage from specification to a deployed prototype.
BSc / MSc · English
Bridge course between the BA and MSc programmes in computer science. Algorithm design and analysis with an emphasis on asymptotic complexity, divide-and-conquer, greedy methods, dynamic programming, graph algorithms, and the basics of NP-completeness. Taught in English.
BSc · Mixed
Introductory BA course developed in 2020 covering computer organisation, operating systems fundamentals, and the practical digital-literacy skills students need across the rest of the degree. Designed for a mixed Icelandic and international audience.
BSc · Mixed
Foundational discrete math course for first-year computer science students. Covers logic and proof techniques, set theory, combinatorics, graphs, recurrences, and an introduction to probability on discrete sample spaces. Taught at the University of Iceland Department of Computer Science with a mixed Icelandic and international cohort.