Fair and Transparent Machine Learning Methods (FTML)
Deploying machine learning models for downstream applications brings with it a wealth of possibilities. However, there is also a non-negligible risk of potential harm if models are not developed carefully.
Data can encode undesired societal biases, which can in turn be
perpetuated by machine learning models if trained on such data.
There may be risks in developing automated solutions for certain
application tasks altogether. Moreover, ML models are often black
boxes whose decisions are not transparent to end-users, creating
imbalances and issues regarding the accountability of models.
Therefore, it is imperative to reflect on the benefits and risks of
ML models, to develop methods to detect and mitigate biases in ML
models, and to create solutions to make the inner workings of
models more transparent. This course focuses on the technical
solutions needed to improve the fairness, accountability and
transparency of machine learning models. As such, it assumes
students have prior knowledge of machine learning.
This course covers the following tentative topic list:
- Statistical notions of fairness and bias
- The intended usage of ML models, e.g. datasheets, model cards
- Learning fair representations, e.g. counterfactual data augmentation, adversarial training, model calibration
- Model interpretability and transparency
- Generating explanations, e.g. post-hoc explainability, generating free-text explanations
- Evaluating model explanations
- Probing representations for bias, e.g. functional testing, subspace probing, generative approaches
MSc Programme in Computer Science
- ML fairness: how to operationalise and measure fairness
- Model bias: how to automatically detect and mitigate ML model biases
- Transparency: interpretability and explainability for ML models
- Develop methods to automatically detect, measure and mitigate biases in ML models
- Develop methods to interpret features deep neural networks have learned
- Develop methods to explain decisions made by ML models
- Transparently document the intended usage of ML models
- Understand methods for bias detection and mitigation, interpretability and explainability
- Plan and carry out fairness and bias analyses on datasets and ML tasks
The format of the class consists of lectures (including guest lectures), presentations by students, and project work.
Selected papers and book chapters. See Absalon when the course is set up.
Knowledge of machine learning (probability theory, linear algebra, classification) and programming is required corresponding to NDAK15007U Machine Learning or NDAB21005U Machine Learning A or similar.
PhD students can register for MSc course by following the same procedure as credit students, see link above.
- 7,5 ECTS
- Type of assessment
Oral examination, during courseWritten assignment, during course
- Type of assessment details
- The exam consists of two parts:
1) A class presentation of an academic paper (oral part)
2) An individual mini project on a topic covered in the course, the findings of which are to be documented in a short report (written part)
The final grade is based on an overall assessment of the assignments and the presentation.
- All aids allowed
- Marking scale
- 7-point grading scale
- Censorship form
- No external censorship
Several internal examiners.
Criteria for exam assessment
See learning outcome
Single subject courses (day)
- Practical exercises
- Project work
- Course number
- 7,5 ECTS
- Programme level
- Full Degree Master
- Block 2
- No limit
The number of seats may be reduced in the late registration period
- Study Board of Mathematics and Computer Science
- Department of Computer Science
- Faculty of Science
- Isabelle Augenstein (10-64786a68717677686c7143676c316e7831676e)
Are you BA- or KA-student?
Courseinformation of students