| Dozent/in |
Ass.-Prof. Christian Frederik Baumgartner |
| Veranstaltungsart |
Vorlesung |
| Code |
FS261662 |
| Semester |
Frühjahrssemester 2026 |
| Durchführender Fachbereich |
Gesundheitswissenschaften |
| Studienstufe |
Master |
| Termin/e |
Do, 19.02.2026, 14:15 - 18:00 Uhr, E.508
Do, 26.02.2026, 14:15 - 18:00 Uhr, E.508
Do, 05.03.2026, 14:15 - 18:00 Uhr, E.508
Do, 12.03.2026, 14:15 - 18:00 Uhr, E.508
Do, 19.03.2026, 14:15 - 18:00 Uhr, HS 4
Do, 26.03.2026, 14:15 - 18:00 Uhr, E.508
Do, 02.04.2026, 14:15 - 18:00 Uhr, E.508
Do, 16.04.2026, 14:15 - 18:00 Uhr, E.508
Do, 23.04.2026, 14:15 - 18:00 Uhr, E.508
Do, 30.04.2026, 14:15 - 18:00 Uhr, E.508
Do, 07.05.2026, 14:15 - 18:00 Uhr, E.508
Do, 21.05.2026, 14:15 - 18:00 Uhr, E.508
Do, 28.05.2026, 14:15 - 18:00 Uhr, E.508
Di, 16.06.2026, 14:00 - 15:00 Uhr, HS 4 (Prüfung) |
| Umfang |
4 Semesterwochenstunden |
| Inhalt |
Advanced machine learning techniques such as large language models (LLMs), vision-language models (VLMs) as well as dimensionality reduction techniques such as t-SNE or UMAP have become foundational tools in data analysis and artificial intelligence. They are increasingly important in healthcare, research, and industry applications. In this course, students will learn to:
• Use state-of-the-art LLMs through common interfaces such as Hugging Face, Google Gemini, and OpenAI.
• Run transformer-based models locally on their laptops and understand practical hardware considerations.
• Apply LLMs to real-world healthcare tasks such as processing large text corpora, screening abstracts, classification, and summarization.
• Work with different model families, including BERT-like models, GPT-style models, and CLIP for multimodal tasks.
• Design effective prompts and recognize common failure modes, biases, and ethical challenges of LLMs.
• Build complete Python workflows that integrate LLMs into data analysis and machine-learning pipelines.
The course combines lectures with extensive hands-on exercises, enabling students to directly apply LLM-based methods to relevant healthcare examples. |
| Lernziele |
• Develop practical skills in using large language models (LLMs) such as BERT and GPT models, as well as vision language models (VLM) such as CLIP or LLaVA.
• Understand and apply advanced dimensionality reduction techniques (e.g., t-SNE and UMAP) to analyze and visualize data.
• Learn to use public API-based workflows (Hugging Face, Google Gemini, OpenAI), and to run deep learning locally on the students’ own laptop.
• Deepen the understanding of Python and build ML pipelines for preprocessing data, running models, and evaluating outputs.
• Learn how to solve common tasks such as screening of large text corpora, classifying text and images, extracting entities and generating summaries.
• Design effective prompts and interaction strategies to reliably elicit useful outputs from LLMs for applied tasks.
• Identify typical failure modes and ethical considerations of LLMs in healthcare, including hallucinations, bias, privacy risks, and responsible use. |
| Voraussetzungen |
The prerequisite for this course is a basic knowledge of Python. Moreover, some initial experience with machine learning is advantageous. |
| Sprache |
Englisch |
| Begrenzung |
Important: The course is limited to 20 participants. The limit is administered via MOODLE according to chronological order and registration. From 2 February 2026, noon, it will be possible to register via MOODLE. As soon as 20 participants are enrolled, the registration window will be closed automatically. |
| Anmeldung |
Moodle: https://elearning.hsm-unilu.ch/course/view.php?id=1009 |
| Leistungsnachweis |
Course assessment. The course will be assessed with three types of learning evaluations:
• Completion of weekly hands-on exercises (20%)
•Written exam on the basics of Python (50%)
• Quality of work and presentation of the final project (30%)
IMPORTANT: In order to earn credits and participate at the exam registration via Uni Portal within the exam registration period is MANDATORY. Further information: www.unilu.ch/en/study/courses-exams-regulations/health-sciences-and-medicine/exams/ |
| Abschlussform / Credits |
weekly exercises, written exam, presentation / 6 Credits
|
| Hinweise |
Teaching methods:
• Lectures: Weekly lectures to introduce and explain core concepts.
• Weekly Hands-On Sessions: Practical exercises and coding sessions to apply concepts learned in lectures.
• Project Work: A final project that can be carried out either alone or in groups of up to 3 people will allow the students to apply the newly develop skills in a creative open-ended setting.
• Presentations: The students will share their project work with the rest of the class in form of a short presentation. |
| Hörer-/innen |
Nein |
| Kontakt |
christian.baumgartner@unilu.ch |
| Material |
The lecture slides, exercise sheets as well as the student presentations will be made available to all students. |
| Literatur |
The course is in part based on the book “Hands-On Large Language Models” by Alammar and Grootendorst. However, students are not required to purchase the book. |