Welcome to the Human-Centered AI Research Seminar (class 2021/22)

Due to the increasing availability of big data, increasing computational power and the great success in data-driven statistical machine learning, the field of Artificial Intelligence as a cross-sectional discipline for all application domains (from smart health to smart farming) is developing into a key discipline of our future society. The Human-centered AI Lab (Holzinger Group) is dedicated to educate the next generation of data science students able to tackle future challenges of human-centered ai and ethcial responsible verifiable machine learning to put the human-in-control of AI and align it with human values, privacy, security and safety.  > Download the Course Syllabus (pdf, 108 kB)

This page is the general information platform for MSc and PhD students and will be adapted to LV 706.996 (5 ECTS, winter term) and LV 706.998 (5 ECTS, summer term) as well as LV 706.997 (2 h, winter term) and LV 706.999 (2h, summer term). The course will be scheduled to the individual needs of each course. Hands on! We speak Python!

The course will be scheduled to the individual needs of each course. Hands on! We speak Python!

“There are two possible outcomes: if your result confirms your set hypothesis, then you have made a measurement. If the result is contrary to your hypothesis … then you have made a novel discovery” (attributed to Enrico Fermi (1901-1954))

Module 01: Introduction to Human-Centered AI

In the first part the students get a rough overview on the concepts of human-centered AI and some currently hot topics from Artificial Intelligence and Machine Learning to get a good common understanding and basis for further research in our areas. Central questions include: What is Human-Centered AI and why is it important? What is integrative machine learning? Why is the health domain complex? What is probabilistic information? What is the difference between autonomous ML and interactive machine learning? What is the human-in-the-loop supposed to do? What is the difference between Causality and Causability? What is ground truth and why do we need grond truth?

Topic 01: The HCAI approach: Towards integrative AI/ML
Topic 02: The complexity of the application area Health Informatics
Topic 03: Probabilistic Information and Probabilistic Learning
Topic 04: Gaussian Processes
Topic 05: Automatic Machine Learning (aML)
Topic 06: Interactive Machine Learning (iML)
Topic 07: Causality, Explainability, Interpretability, Causability
Topic 08: The #KandinskyPatterns Exploration Environment

Lecture slides 2×2 (pdf, 9,996 kB):
Lecture slides full size (pdf, 6,121 kB):

Note: The slides provided here are for printing and reading, the slides shown will be different from a didactial point of view.

Module 02: The Fundamentals: Theory of Science

In the second part the students get a rough overview on the basic concepts of science (Wissenschaftstheorie). We start with answering the question of what science is and why we should contribute to the international scientific community. We are engineers, so do we need the scientific methodological approach? We see science as the field to provide explanations and to make predictions, so it is highly relevant for any data science student!

Note: Particularly the field of AI/machine learning needs a concerted effort of many people in an international concerted effort *). This also ensures quality control and most of all ethical control.

Topic 01: Motivation: Why should we contribute to the international scientific community?
Topic 02: What is Science? What is Engieering?
Topic 03: Basics of the Theory of Science (Wissenschaftstheorie)
Topic 04: Hypothetico-Deductive Method (the scientific method)
Topic 05: Occam’s Razor
Topic 06: Reichenbach’s Principle
Topic 07: Experiments in Machine Learning
Topic 08: #KANDINSKYPatterns – our “Swiss Knife” for the study of explainable AI
Topic 09: Sample Questions and Conclusion

Lecture slides 2×2 (pdf, 9,996 kB):
Lecture slides full size (pdf, 6,121 kB):

Note: The slides provided here are for printing and reading, the slides shown will be different from a didactial point of view.

Module 03 – The Mechanics: Management of Research

In Module 2 we discussed questions including “What is science?”, “Why contributing to the international scientific community?” and some methodological issues of the Theory of Science (“Wissenschaftstheorie”). Now, in Module 3 we get hands-on and discuss questions including
“How to contribute to the international scientific community?” and learn the basic mechanics of science, the “know‐how”. Of course always with a focus on explainable AI and ethical responsible machine learning.

Topic 01: Introduction and Overview: Successful Management of Research & Development
Topic 02: Workflows for a Research Group
Topic 03: Measurable Output: Publications (What is a “paper”?)
Topic 04: How do I read a paper?
Topic 05: How to find (relevant) papers?
Topic 06: How to manage papers?
Topic 07: Publication targets
Topic 08: How to write a paper?

Lecture slides 2×2 (pdf, 9,996 kB):
Lecture slides full size (pdf, 6,121 kB):

Note: The slides provided here are for printing and reading, the slides shown will be different from a didactial point of view.

Module 04: Local Specifics

In the last part of our research seminar, the students get some practical hints on the local processes of how to do a Bachelor Thesis, a Master’s Theses and partiucularly a PhD thesis.

Fast Track:
1) Watch the Students Welcome Video https://goo.gl/GoeBek
2) Read the introduction paper https://www.mdpi.com/2504-4990/1/1/1
3) Read the research and teaching statements, availabe at https://www.aholzinger.at/
4a) PhD-students develop carefully your PhD-proposal PhD-Proposal-HOLZINGER-Group-2017
4b) Latex-Template Masterthesis for Students of the Holzinger Group
4c) Bachelor Students just enroll to course 706.170 at TUGOnline and select Group Holzinger
5) Deans Pages of the Computer Science faculty: https://www.tugraz.at/fakultaeten/infbio/faculty/team/deans-office

Topic 01: Bachelor – your first academic and/or practical computer science work
Topic 02: Master – your specialization in computer science
Topic 03: PhD – your rocket to science and industry
Topic 04: Specific PhD checklists (from the first discussion to the doctoral defense (Rigorosum))
Topic 05: Everything starts with the PhD proposal: (a) Written document, (b) Presentation, (c) Discussion
Topic 06: Risk mitigation and fallback solutions

Lecture slides 2×2 (pdf, 9,996 kB):
Lecture slides full size (pdf, 6,121 kB):

Note: The slides provided here are for printing and reading, the slides shown will be different from a didactial point of view.