Bayesian Optimization in Quantitative Finance

This repository contains the code and results accompanying the MSc thesis
"Advancing Bayesian Optimization in Quantitative Finance" (University College London, 2025).

Contents

• 01_framework_comparison.ipynb — Comparison of Hyperopt and Optuna on a synthetic benchmark function
• 02_batch_bo.ipynb — Sequential vs. Batch Bayesian Optimization in the DRACUS backtesting environment
• 03_results_oos_robustness.ipynb — Out-of-sample validation, robustness and sensitivity analysis
• MSc_Thesis_Bayesian_Optimization_Fuchs.pdf — Full thesis

Note: The DRACUS backtesting system referenced in the thesis is proprietary and not included in this repository.
The notebooks illustrate its usage conceptually, but the implementation itself is not publicly available.

Usage

The notebooks can be opened and run with Jupyter.
A standard Python 3.10+ environment with the following packages is required:
numpy, pandas, matplotlib, optuna, hyperopt.

License

This project is released under the MIT License.