Adversarial learning aims to identify weaknesses in machine learning models. The goal is to identify potential problems that cannot be found using traditional evaluation using test sets. It has been used successfully in a wide range of applications, typically focused on a specific model or domain. A direction of adversarial learning aims to identify examples that could break or improve the training of the model if that example would be added to the training. In this project, we will use adversarial learning to identify potential problems in the training process of a model, aiming to find a harmful bias in the data, such as under-represented groups. If these biases are not identified in data and models, a machine learning model will not produce correct predictions for all cases.
Duration and Type
- The summer scholarship is over 12 weeks between S2 2020 and S1 2021
- As project preferable 1 semester, but can be also done in 2 semesters
- Honours project, other postgraduate project (MSc, MProfStuds, …), summer scholarship
- Basic maths, CS, and machine learning skills
- Programming in Java or Python