Dimitri Malandreniotis
14 Jan 2026
A Collaboration with UCL to offer two advanced machine learning projects for postgraduate research
Autonomous-Fox Laboratories is pleased to participate as an industry partner in the UCL Industry Exchange Network (IXN) Innovations Programme, in collaboration with UCL’s Machine Learning community.
Through IXN, postgraduate students from UCL’s MSc Machine Learning, MSc Data Science and Machine Learning (DSML) and MSc Computational Statistics and Machine Learning (CSML) programmes will work directly with Autonomous-Fox Laboratories on research-led industry projects addressing open problems in financial AI and time-series modelling.
This collaboration forms part of AF Labs’ broader research programme in discrete representations, structured probabilistic modelling and geometric deep learning for complex networks of interdependent time-series.
The full presentation:
Time-Series Tokenisation: Bringing Financial Market Modelling to the Forefront of Modern AI
This project investigates novel methods for tokenising financial time-series, transforming noisy continuous price dynamics into discrete latent state sequences that encode semantically meaningful aspects of market behaviour such as regime, momentum and volatility structure. Students will explore vector-quantisation-based and alternative discretisation strategies and evaluate their downstream utility for forecasting and regime identification using modern sequence models.
Market Modelling with Geometric Deep Learning on Dynamic Graphs
This project frames the financial market as a dynamic graph system, where assets are nodes and evolving relationships are learned endogenously via attention-based graph neural networks and graph transformers. Students will investigate causal, sparse and regime-adaptive cross-asset modelling techniques for short-horizon forecasting using tokenised market data.
We look forward to welcoming the selected students in June.