Training Week on Blockchains in Digital Finance

This course focuses on applying statistical modeling, machine learning (ML), and artificial intelligence (AI) methods—particularly large language models (LLMs)—to blockchain-based financial systems. 

Students will explore the use of these technologies in: 

• Predictive modeling of crypto assets, 

• Anomaly detection in decentralized finance (DeFi), 

• Transaction graph analytics, 

• Smart contract evaluation, 

• Blockchain-based risk assessment. 

Emphasizing data-driven research, the course makes extensive use of real-world blockchain datasets from sources such as Quantinar, Dune Analytics, Etherscan, and Flipside Crypto. Through lectures and applied labs, students will formulate, implement, and evaluate models that capture the complexity of digital finance systems built on distributed ledger technologies. 

The course positions students at the forefront of quantitative blockchain analytics, with practical exposure to ML frameworks (e.g., PyTorch, scikit-learn), financial time series analysis, and the responsible use of LLMs for interpretability, auditing, and compliance in DeFi. 

Modules and Credits

This training week is part of the MSCA Digital Finance Doctoral Program:

• 30 March - 02 April: Blockchains in Digital Finance (4 ECTS)

• 03 April: HE framework and research project management (1 ECTS)

Prior Knowledge

To successfully participate in this course, students are expected to have foundational knowledge in the following areas: 

• Statistics & Econometrics 

  Familiarity with linear regression, maximum likelihood estimation, GARCH models, and hypothesis testing. 

• Machine Learning 

  Understanding of core supervised and unsupervised algorithms (e.g., decision trees, clustering, neural networks), including basic concepts of deep learning and transformers. 

• Mathematics 

  Solid grasp of linear algebra (eigenvalues, matrix operations), probability theory (distributions, expectations), and convex optimization. 

• Programming 

  Proficiency in Python, with experience using libraries such as NumPy, pandas, scikit-learn, PyTorch, and Hugging Face Transformers. 

• Finance 

  Basic understanding of financial markets, asset pricing, and blockchain-specific financial mechanisms (e.g., tokenomics, AMMs, DeFi protocols). 

• Blockchain Fundamentals 

  Familiarity with blockchain architecture, smart contracts (e.g., Solidity), Ethereum, and key tools such as Web3.py and Etherscan. 

Time Schedule

Speakers

• Adrian Costea Bucharest University of Economic Studies

• Antoaneta Amza National Bank of Romania

• Cătălin Dumitrescu Stratum Finance

• Claudia Petrescu The Research Institute for Quality of Life, Romanian Academy

• Claudia Voicilă National Bank of Romania

• Dan Gabriel Anghel Bucharest University of Economic Studies

• Daniel Traian Pele IDA Institute Digital Assets, Bucharest University of Economic Studies; Institute for Economic Forecasting, Romanian Academy, Romania

• Joerg Osterrieder Coordinator MSCA Digital Finance

• Oros Alexandra Omnisource Technologies; PMI Romania

• Owen Chaffard Cardo AI; University of Kaiserslautern-Landau

• Piotr Wójcik University of Warsaw, Poland

• Radu Lupu IDA Institute Digital Assets, Bucharest University of Economic Studies; Institute for Economic Forecasting, Romanian Academy, Romania

• Raluca Micu TOKEN Financial Technologies Romania

• Raul Cristian Bag Institute of Digital Assets, Bucharest University of Economic Studies

• Vlad Bolovăneanu Bucharest University of Economic Studies

• Wolfgang Karl Härdle IDA Institute Digital Assets, Bucharest University of Economic Studies, Romania

Accommodation Options and Transport Information

Acknowledgment

This training week is supported by G-Research, and the MSCA Digital Finance network acknowledges this funding support.