Aims and Fit of Module

This module introduces emerging deep learning models and their applications in financial markets. It aims to equip students with a comprehensive skillset, enabling them to demonstrate critical understanding of widely employed models and apply hands-on techniques to model real-world financial data of different modalities (e.g. time series, texts) . Students will develop a deep understanding of fundamental concepts, including neural networks, deep feed-forward neural networks, recurrent neural networks, Long Short-Term Memory, transformers, transfer learning, generative model, and topics of large language model (LLM). The module will also cover techniques for modelling financial data using cross-validation and ensemble methods, as well as state-of-the-art unsupervised learning methods like autoencoders, applied to company clustering and asset pricing. Additionally, students will explore the latest machine learning interpretability techniques and evaluate feature importance using various methods

Learning outcomes

A. Apply and demonstrate a critical understanding of algorithms and state-of-the-art neural network architectures, such as deep feed-forward neural networks, recurrent neural networks, and Long Short-Term Memory, for financial and macroeconomic data modelling. B. Apply and demonstrate a critical understanding of state-of-the-art textual analysis models and related sentiment analysis methodologies. C. Demonstrate critical understanding of various types of financial data, their underlying characteristics and challenges, and develop potential solutions for modelling mixed-frequency financial and macroeconomic data, D. Apply techniques for modelling financial data using cross-validation and ensemble methods. E. Demonstrate a critical understanding of state-of-the-art unsupervised learning methods, such as autoencoders, and their application in company clustering and asset pricing. F. Evaluate and apply state-of-the-art machine learning interpretability techniques to assess feature importance through different methods.

Method of teaching and learning

This module is delivered through formal lectures and tutorials.