It is well-known that traditional linear statistical methods are often outperformed by current non-linear (deep) machine learning methods that have shown exceptional high-precision in automated prediction models. However, actual applications of these better performing models remain rare, as the non-linear machine learning techniques (such as the currently popular deep learning methods with many kinds of processing layers) are presumed black-box. Explainable artificial intelligence refers to machine learning techniques, or generally methods in artificial intelligence, for which the underlying decision logic and outcomes can be explained. My research compares current techniques and explores novel approaches to make high-precision automated decision-making explainable to domain experts. The Tutkijat Maailmalle grant would enable me to conduct a research visit at the Rensselaer Polytechnic Institute (i.e., the fourth best university for information technology, see niche.com/colleges/search/best-colleges-for-information-technology ) that employs world-leading scholars in knowledge-representation and reasoning. This would be extremely beneficial for my research, especially for adding semantics to the explanations.

2021

2022