PhD candidate
Research group | BioAI
Main supervisor | Kai Olav Ellefsen
Co-supervisor | Mikkel Lepperød & Marianne Fyhn
Affiliation | Department of Biosciences, UiO
Contact | mateusz.wasiluk@ibv.uio.no
Short bio
My interest in Artificial Intelligence began during my undergraduate studies in Applied Physics at the Warsaw University of Technology. I was fascinated by the concept of combining machine learning algorithms and basic research in nuclear physics. As a result, I developed several models, for problems like elementary particle identification and event selection, in cooperation with big, European accelerator facilities - CERN and FAIR. To further broaden my understanding of AI and algorithmics in general, I decided to pursue a second Master of Science degree, this time in the field of Computer Science and Engineering at the Technical University of Denmark. However, after just three semesters another opportunity arised and I left the plains of mainland Europe to engage in the CompSci PhD programme.
Research interests and hobbies
My scientific interests are strongly connected to interdisciplinarity, either between physics and computer science, or biology and AI. I believe that drawing inspiration from different fields is the key to the full understanding of the world we live in.
When I’m not “doing science” I enjoy hiking, singing and dancing. I’m also interested in Nordic culture, polar research and everything related to cold!
CompSci project
Bio-inspired methods for continual learning in deep neural networks
In the past decade, the field of Artificial Intelligence (AI), and more precisely, Machine Learning (ML) has experienced enormous growth in terms of scientific interest and the number of industrial applications. While self-improving algorithms have already become widespread in the modern world, largely due to their ability to specialise, there are still scenarios in which they fail to meet the requirements of a more challenging training process. One such scenario is continual or lifelong learning, in which a model experiences a sequential stream of training episodes and needs to adjust to novel input distributions or tasks while still performing well on the previous ones. Despite such an ability being a standard property of biological cognitive systems (like humans), the lack of it in artificial neural networks (ANNs) is still an open problem. It manifests itself mainly as the phenomenon of catastrophic forgetting (CF) in which a new set of patterns suddenly and completely erases a network’s previous knowledge. Most of the current methods of overcoming this issue either strongly specialise in a particular area of applications (like image recognition) or are too computationally expensive to be used efficiently (e.g. by storing all the incoming data and regularly retraining the whole model). The main goal of this project is to draw from mechanisms that enable continual learning in biological agents and use them as an inspiration for novel methods of preventing catastrophic forgetting in artificial neural networks.
Publications
CompSci publications
None yet.
Previous publications
None.