Quantitative Finance with Machine Learning

Overview

This repository demonstrates the application of advanced Machine Learning (ML) techniques in the field of Quantitative Finance. It showcases a variety of ML-driven approaches to key financial tasks, aiming to solve complex problems while improving decision-making processes in areas like portfolio optimisation and financial forecasting.

The goal of this repository is twofold:

• To explore state-of-the-art ML algorithms in Quantitative Finance.
• To serve as a demonstration of my expertise for future career opportunities in the finance and data science sectors.

Current Work

For now, this repository currently includes two main areas of focus:

• Portfolio Optimisation:
  • Implementations include classic optimisation algorithms as well as Monte Carlo simulations to optimise portfolio returns.
• Financial Forecasting:
  • This section focuses on time series forecasting for stock prices, using multivariate, feature engineering and with easy adaptability to predict other financial instruments such as indices, commodities, ETFs, etc.
  • Models used include cutting-edge approaches like Long Short-Term Memory (LSTM) networks, Transformer models, and state-of-the-art transfer learning architectures, such as Chronos, a breakthrough model introduced by Amazon in 2024, leveraging Large Language Models (LLMs) for time series predictions.

Future Updates

Several planned improvements and additions to this repository include:

• Advanced Optimisation Techniques:

  • Ongoing work on incorporating Reinforcement Learning and Bayesian Optimisation to further enhance portfolio optimisation methods.

• Expanding ML Use Cases:

  • Introducing new machine learning tasks that address other crucial problems in Quantitative Finance.

• Quantum Computing Applications:

  • Integrating Quantum Computing methods to tackle the presented tasks with higher efficiency. This is an area where I have substantial interest and experience. (For more, I encourage you to explore my other repositories on Quantum Computing.)

Requirements

• gpytorch >=1.12
• botorch >= 0.11
• chronos
• matplotlib >=3.9
• seaborn >=3.9
• mlflow >=2.16
• optuna