Digital Finance - Reaching New Frontiers
DIGITAL is an Industrial Doctoral Network established within the Marie Sklodowska-Curie Actions which aims at training highly skilled doctoral candidates in Digital Finance.
Fiche signalétique
- Départements participants Gestion
- Institut(s) Institute for Applied Data Science & Finance
- Unité(s) de recherche Applied AI Research & Solutions
- Organisation d'encouragement Europäische Union
- Durée (prévue) 01.01.2024 - 31.12.2027
- Direction du projet Prof. Dr. Branka Hadji Misheva
-
Équipe du projet
Prof. Dr. Lucia Gomez Teijeiro
Prof. Dr. Adam Andrzej Kurpisz
Rebecca Di Francesco
Julius Kooistra - Mots-clés Digital finance, artificial intelligence, explainable AI, time-series forecasting, sustainable finance, ESG, portfolio optimisation, secondments, ECB, ING, financial regulation, trustworthy AI
Situation
The rapid transformation of financial services through artificial intelligence, machine learning, and data-driven decision systems has created a strong need for highly trained researchers capable of bridging methodological innovation and real-world financial applications. In particular, explainable AI (XAI) and sustainable finance have become central requirements for ensuring transparency, regulatory compliance, and trust in automated financial decision-making. Within this context, BFH plays a key role in the consortium through leading two work packages and supervising two doctoral candidates. These candidates work at the intersection of: - Explainable AI for financial time series, focusing on robustness under non-stationarity, regime shifts, and model uncertainty - Sustainable and responsible finance, including portfolio decarbonisation, ESG constraints, and greenwashing detection Their work directly addresses key project’s objectives on deployment-ready explainability and stakeholder-aware AI systems for finance.
Approche
The two BFH-supervised doctoral candidates work in a highly collaborative, internationally networked environment combining academic research with industry exposure through secondments at key partners, including the European Central Bank (ECB) and ING. These placements enable validation of research in real financial settings, supporting MSCA goals of industry integration and transferable skills development.
Résultat
- Development of explainable AI methods for financial time series, including systematic analysis of SHAP, LIME and gradient-based methods, identification of limitations under unique stylized facts, and exploration of new approaches for improved interpretability in forecasting models. - Advancement of sustainable finance models, including portfolio decarbonisation under tracking-error constraints. - Emerging prototypes for explainable AI in finance, including benchmarking and evaluation frameworks for time-series models and implementations of ESG- and decarbonisation-aware portfolio optimisation models. - Development of reproducible pipelines for XAI in financial forecasting and dashboards for sustainable portfolio construction. - Strong industry collaboration through secondments at the European Central Bank, focusing on stress testing and credit impairment forecasting, and ING, focusing on generative AI for ESG reporting, decarbonisation analysis and client engagement in sustainable finance - Regular research seminars within WP3 supporting methodological exchange on explainable AI, trustworthy AI, and sustainable finance, enabling continuous feedback across doctoral candidates and partner institutions - BFH-led training activities in explainable AI, including structured doctoral-level modules and seminar-based training on XAI for finance.
Perspectives
The next phase of the project focuses on converting research outputs into higher-impact scientific publications, validated prototypes, and policy-relevant frameworks.
