Where I am asking questions that have not been answered yet.
Research here is separated into three categories — what has been published, what I am actively working on, and what I intend to do. They are not the same thing, and I keep them apart.
Published
No peer-reviewed publications yet. Work in progress is listed below under ongoing and proposed research.
Ongoing
Machine Learning and Large Language Models for Predictive Modelling of African Frontier Market Returns
An ongoing investigation into whether frontier-market returns carry structure that conventional models miss, and whether large language models can contribute signal beyond standard quantitative factors. The work has an applied dimension: the model-building approach has already been exercised in the Godtrader market-research platform, where machine-learning components support signal generation and strategy evaluation.
Proposed
Predictive Attendance Forecasting for Large Distributed Workforces
Applying the forecasting techniques used in market modelling to organisational attendance data — using historical attendance patterns to anticipate shortfall and coverage risk across distributed divisions before it becomes an operational problem.
Evaluation Frameworks for LLM-Assisted Financial Analysis
A proposal for a reproducible way to assess where large language models add genuine analytical value in financial modelling, and where they introduce unsupported reasoning into a decision pipeline.
Primary research interest
Machine Learning and Large Language Models for Predictive Modelling of African Frontier Market Returns.