Artificial Intelligence in Finance
The AI for Finance Lab at the UCL Institute of Finance and Technology is dedicated to advancing research in the application of AI and quantitative methods for financial data analysis. The lab's work includes developing new approaches for liquidity risk estimation, risk profiling, and risk management. These research activities have practical implications, helping financial professionals to better understand and predict market behaviours. Current research collaborations involve the Saudi Central Bank, Santander UK, Consob, the UZH Blockchain Center and the DLT Science Foundation. The lab also offers industry services ranging from AI validation, benchmarking AI implementations to verify they are performing correctly and responsibly, to analytics design, supporting organizations in developing systems that can handle complex financial data effectively. The lab aims to bridge the gap between academic research and industry practice, providing insights that are both innovative and applicable in real-world financial contexts.
- AI Validation
- Ensuring Reliability and Accuracy: The lab focuses on verifying that AI models and systems used in finance perform as intended, are reliable, and produce accurate results. This involves rigorous testing against various scenarios to ensure robustness.
- Ethical and Regulatory Compliance: The lab also ensures that AI applications meet ethical standards and regulatory requirements, crucial in the highly regulated finance sector.
- AI Benchmarking
- Performance Measurement: The lab conducts benchmarking to measure the performance of AI systems against industry standards. This includes comparing AI models in terms of accuracy, speed, and efficiency in handling financial data.
- Identifying Best Practices: Through benchmarking, the lab identifies best practices and aims to set performance benchmarks for AI applications in finance, guiding the development and improvement of AI systems.
- Financial Data Analytics Design
- Developing Advanced Analytical Tools: The lab designs sophisticated tools for analyzing financial data, helping to uncover market trends, risks, and opportunities.
- Customization and Innovation: Focus on customizing analytics tools to suit specific financial contexts and continuously innovate to handle the complexities and evolving nature of financial markets.
Recent publications
S. Wattanawongwan, C. Mues, R. Okhrati, T. Choudhry, & M. C. So. Modelling credit card exposure at default using vine copula quantile regression. European Journal of Operational Research (2023).
C. B. Hamill, R. Khraishi, S. Gherghel, J. Lawrence, S. Mercuri, R. Okhrati, & G. A. Cowan. Agent-based Modelling of Credit Card Promotions. arXiv preprint arXiv:2311.01901 (2023).
R. Khraishi & R. Okhrati. Simple Noisy Environment Augmentation for Reinforcement Learning. arXiv preprint arXiv:2305.02882 (2023).
C. Campajola, F. Lillo, P. Mazzarisi, and D. Tantari, "On the Equivalence between the Kinetic Ising Model and Discrete Autoregressive Processes," Journal of Statistical Mechanics: Theory and Experiment, 2021(3), 033412, 2021.
C. Campajola, F. Lillo, P. Mazzarisi, and D. Tantari, "On the Equivalence between the Kinetic Ising Model and Discrete Autoregressive Processes," Journal of Statistical Mechanics: Theory and Experiment, 2021(3), 033412, 2021.
C. Campajola, F. Lillo, and D. Tantari, "Unveiling the Relation Between Herding and Liquidity with Trader Lead-Lag Networks," Quantitative Finance, 20(11), 1765-1778, 2020.
P. Mazzarisi, S. Zaoli, C. Campajola, and F. Lillo, "Tail Granger Causalities and Where to Find Them: Extreme Risk Spillovers vs Spurious Linkages," Journal of Economic Dynamics and Control, 121 (104022), 2020.
C. Campajola, F. Lillo, and D. Tantari, "Inference of the Kinetic Ising Model with Heterogeneous Missing Data," Physical Review E, 99(6), 062138, 2019.