MSc Financial Data Science

Asset prices in financial markets go up and down in accordance with how markets digest the flow of information. To understand the causal relation between information flow and price movements, it is necessary to model market information and use this to infer the price dynamics. In this way, market dynamics can be replicated artificially on a computer. This module explains the powerful process of artificially generating realistic market models. The module begins with elements of probability theory. We will then learn the idea of conditional expectation and the Bayes formula, which gives the optimal inference under uncertainty. The meaning of the Bayes formula will be explained, leading to the understanding of what is meant by “intelligence”. The module then covers the basics of stochastic process (specifically, the Brownian motion) and calculus (specifically, the Ito calculus), sufficient to follow the contents of the module. Then simple models for flows of information in financial markets will be introduced, and by use of the Bayes formula the associated price dynamics will be derived. The module concludes with a brief application to the asset valuation problems in financial markets (such as options or other derivatives).

Mathematics underpinning real-world uncertain events has become indispensable in many applications, including in particular financial markets. This module will begin with the introduction to probability theory and stochastic processes, with an emphasis on the Ito calculus for treating functions of Brownian motion. Such functions are commonly used in financial markets to model asset price dynamics, required for the valuation of financial contracts. The module then discusses structures of financial markets, with an emphasis on the equity market. Several of the standard and exotic contingent claims will be introduced, and the need for mathematical models for the valuation and risk management of these products will be explained. The pricing of a standard call option will then be worked out in a single-period binomial model, for which option price will be worked out in two ways: first using the portfolio replication and no arbitrage argument, and second using the risk-neutral expectation argument. The model is then extended into multi-period binomial tree model, leading to the Cox-Ross-Rubinstein option pricing formula. Finally, a continuous-time geometric Brownian motion model, originally introduced by Samuelson, will be considered, and used to deduce the famous Black-Scholes option pricing formula. This can be applied for the purpose of both pricing, as well as risk-management purposes, which will be demonstrated by working out the hedging strategy. The meaning of the pricing formula, and how it can be used in practical investment banking context, will be explained.

Mathematics is a key tool in Data Science. This module is designed to introduce students to the foundational mathematical techniques that are required to support future data science modules.

This module provides a solid foundation in Python programming relevant for data science. It introduces students to core programming concepts, essential Python libraries, and practical coding techniques widely used in data analysis and machine learning. By the end of the module, students will be confident in writing Python programs for solving real-world data problems, handling data, performing analysis, and creating visualisations.

Bayesian Statistics is the branch of statistics that relies on subjective probability to create a wide range of statistical models. This module introduces Bayesian methodology and guides students to use prior to posterior analysis for modelling realistic problems. This module then tackles more difficult topics such as Bayesian point estimates, model selection and linear regression.

The module has the aim of introducing students to core concepts in the blockchain technology. Students will become familiar with applications of cryptography to financial transactions and the technicalities behind cryptocurrencies and their workings. Using case studies and simulations, students will be exposed to the use of software wallets and engage in blockchain and cryptocurrency related transactions.

The module provides for coverage of a variety of statistical methods, including descriptive statistics and validating formulated hypotheses, as well as predictive analytics. The computational foundations and methods of importance to data science are also covered, along with consideration for relevant supporting software and tools.

This module introduces programming in Python for data science, with a focus on data pre-processing, data mining and analysis, machine learning and deep learning. Besides the practical hands-on experience with writing code, this course also covers the theoretical background on different data analysis techniques and machine learning approaches. The goal is to develop an understanding of how information can be extracted from data and how this information can be further used to make predictions, but importantly how this is done practically in terms of writing clear and transparent source code. Using real-world data sets and illustrative examples, this course will help to develop a theoretical understanding of data science as well as practical experience by developing useful software tools. Many of the techniques acquired through this module are likely to be of potential use in the dissertation project.

The dissertation consists of a written report of around 50 pages completed by the student towards the end of their programme of study. The report is based on a major piece of work that involves applying material encountered in the taught component of the programme and extending that knowledge with the student's contribution, under the guidance of a supervisor. The work for the dissertation and the writing up begins approximately May/June, continues through the Summer and the dissertation report is submitted in late Summer, normally in the second week of September. The work may, but need not, involve original research. It may instead consist of a substantial literature survey on a specific topic.

Machine Learning for Data Science incorporates a wide range of machine learning algorithms and data mining techniques, which can be applied to real-world problems and datasets with various characteristics to generate new insights and understanding. Through treatment of the principles and fundamental requirements for machine learning, example applications, and related exercises, this module will offer coverage of a range of contemporarily important and emergent machine learning algorithms. The module will provide for the means to critically evaluate, extend, and apply, appropriate techniques to datasets exemplifying specific characteristics in order to derive suitable and defensible results.

Data science is the study of data to extract meaningful and actionable insights at all levels of society such as dynamical systems and social media networks. This module introduces the role of data in society and provides students with the underpinning mathematics that drives data methodology and algorithms. This module then covers wide-ranging topics with a focus on the Surrey brand of data, as research into data is part of the School's research agenda

Time series are a collection of observations taken over time. This covers a great deal of situations such as stock markets, rainfall or even goals scored by a sports team. Features of the time series will lead to an appropriate choice of model. These models will be validated, and then can be used to forecast the future. Despite the modest pre-requisites of Level 4 Probability and Statistics (MAT1033), students will gain resourcefulness and resilience through learning mathematical proofs as well as gain digital capabilities through using R to conduct analyses of data sets and writing a report.

This module introduces students to the research skills required to engage in data science projects in both industry and academia, whilst also covering the relevant ethical and security considerations when designing and implementing data driven projects. Areas of specific concern for ethics and security in machine learning and statistical analysis are highlighted. Good practice in the responsible and effective use of generative AI for research development and academic writing will be demonstrated and critically discussed. This module serves as an important initial module towards planning and executing the dissertation project. It also provides a platform to facilitate team building skills and make the student aware of the challenges of working effectively with other people embracing equality, diversity and inclusion.