Machine learning (ML) sits at the intersection of several thriving disciplines including computational statistics, pattern recognition, and data science. In this course, we focus on the use of ML in analysis of empirical distributions of economic variables, with a key emphasis on developing financial theory. We present machine learning as a higher-dimensional extension to various numerical concepts in quantitative finance (including classical econometric models that rely on simplistic statistical assumptions, and dynamic programming); demonstrating how these techniques complement the traditional methods. As a practical discipline, students will learn to understand how and when to effectively make use of a range of ML techniques and knowledge representations (e.g., classification, clustering, regression and association) for modelling, inferring and predicting phenomenon observed in real-world datasets, using Python and associated industry standard libraries.

One two-hour workshop per week for 10 weeks

6 1-and-a-half-hour computer labs

None

The aim of this course is for the students to gain a comprehensive understanding of the use of machine learning (ML) techniques to discover economic and financial theories, with a particular emphasis on unique challenges of analysis of financial data. Students will learn to efficiently apply these methodologies, determining appropriate algorithm, optimisation technique applicable for the given data set and problem domain, under a standard data mining taxonomy: Objective Specification, Data Curation and Exploration, Data Cleaning, Feature Selection, Model Selection, Parameter Turning, Evaluation. On completion of this course, students should be able to analyse large volumes of financial data programmatically and have a practical experience working with standard ML libraries and packages in Python.

By the end of this course students will be able to:

1. Analyse and formulate objective of ML problem and implement solutions using ML libraries and packages in Python.

2. Apply different optimisation algorithms and methods through appropriate ML applications.

3. Critically analyse different financial datasets while keeping in mind the computational requirements and performance limitations of different ML algorithms.

4. Develop ML applications while programming collaboratively in small groups.

Students must submit at least 75% by weight of the components (including examinations) of the course's summative assessment.