Boundaries of Explainable AI in Financial Time Series
- stephanieblum7
- Jun 5, 2025
- 1 min read
Studies on Explainable AI (XAI) have greatly increased in number - but there are issues. In a research seminar at Bern University of Applied Sciences (Business School), PhD candidate Jens Reil presented the first results of his ongoing research into explainable artificial intelligence (XAI) for financial time series.
Jens's project focuses on how current XAI methods perform when applied to time-dependent data. This includes common techniques such as SHAP and LIME, which are widely used to make black-box models more transparent.
In his structured literature review, Jens highlights that while the number of XAI studies is growing fast, many rely on outdated assumptions. In the financial domain especially, researchers tend to apply XAI tools without questioning whether the explanations are actually accurate or meaningful.
He highlights a key issue: many XAI methods break the time sequence of the data. As a result, their explanations can lead to contradictory or non-comparable results. To address this, Jens calls for new, tailored approaches that respect time dependencies and avoid synthetic distortions. His goal is to help build XAI tools that work more reliably in complex, real-world financial systems.
The project is funded by the European Union’s Horizon Europe programme under the Marie Skłodowska-Curie Actions.
nice update 👍 btw i usually use an instagram video download tool to save clips for offline viewing
The content was really very interesting. I am really thankful to you for providing this unique information You have a good point here! I totally agree with what you have said!! Thanks for sharing your views. Skeleton Jacket
Extremely go through article. I have never seen such beautiful article. Learn many things from this I hope you continue to have high-quality articles like this to share with everyone! Michael Jackson Red Jacket
I appreciate how clearly this article explains the balance between accuracy and interpretability in financial models. It’s not easy to simplify such a technical topic. While taking a quick pause after reading, I sometimes relax with GeoDash for a few minutes.
While browsing a few different fashion websites earlier today, I noticed that many hoodies look almost identical. That’s why on Vanson Jackets the Elwood Strawberry Print Hoodie felt refreshing because the design adds character without making the hoodie feel over the top.