Statistics with Data Analytics MSc

Programme structure

Your studies on the course will cover the modules listed below.

Compulsory modules

Quantitative Data Analysis

The aim of the module is to develop knowledge and skills of the quantitative data analysis methods that underpin data science. Content covers a practical understanding of core methods in data science application and research (e.g., bivariate and multivariate methods, regression and graphical models). A focus is also placed on learning to evaluate the strengths and weaknesses of methods alongside an understanding of how and when to use or combine methods.

Research Methods and Case Studies

The aims of this module are to develop students’ knowledge and critical awareness of a variety of research methods; encourage students to develop critical thinking skills and transferable skills appropriate to their discipline, enable students to develop an understanding of the current needs of industry and commerce, prepare students for their dissertation.

Computer Intensive Statistical Methods

In this course, you will learn six computer intensive methods by R. These six methods are:

• Basic Monte-Carlo simulation methods
• Simulating from specified distributions which include inverse method and rejection sampling;
• Aggregation and compression technique based on the “divide-and-conquer” strategy to cope with massive data in regression analysis;
• Bootstrap which is used in several contexts, most commonly to provide a measure of accuracy of a parameter estimate or of a given statistical learning method;
• Randomization testing and Monte Carlo approach such as t-test without the normal assumption;
• Monte-Carlo integration such as importance sampling for numerical integration with good accuracy.

Modern Regression and Classification

This course aims to equip students with the knowledge and ability to use modern regression and classification methods with Big Data, particularly apply a range of models and tools to variable selection and model selection.

• Demonstrate a working knowledge of the theory of Lasso and its different versions.
• Develop a working knowledge of the research concepts involved in p>n.
• Select a classification for a specific application and interpret the output of the statistical analysis.
• Fit Lasso-based models to data from a variety of applied domains.
• Implement variable selection and statistical classification, using a software package.
• Interpret and visualise the output of the statistical inference.

Data Visualisation

The aim of the module is to develop the reflective and practical understanding necessary to visually present insight drawn from large heterogeneous data sets (e.g., to decision-makers). Content will provide an understanding of human visual perception, data visualisation methods and techniques, dashboard and infographic design. The role of interactivity within the visualization process will be explored and an emphasis placed on visual storytelling and narrative development.

Big Data Analytics

The aim of the module is to develop the reflective and practical understanding necessary to extract value and insight from large heterogeneous data sets. Focus is placed on the analytic methods/techniques/algorithms for generating value and insight from the (real-time) processing of heterogeneous data. Content will cover approaches to data mining alongside machine learning techniques (e.g., clustering, regression, support vector machines, boosting, decision trees and neural networks).

Time Series Modelling

This course aims to equip students with the ability to employ different methods for modelling and forecasting time series data, in particular in the context of financial data and forecasting financial risk, and to apply a range of models and tools to make financial decisions such as risk assessment.

Network Models

This courses aims to equip students with the ability to employ different methods for modelling networks, and to apply a range of network models and tools to a number of applied domains. The contenst include

• Demonstrate a working knowledge of the theory of network models and of the research concepts involved in inferring and summarising networks.
• Select a network model for a specific application.
• Fit network models to data from a variety of applied domains.
• Implement statistical network models, using a software package.
• Interpret and visualise the output of the statistical inference.

Statistics with Data Analytics Dissertation

Towards the end of the Spring Term, students will choose a topic for an individual research project, which will lead to the preparation and submission of an MSc dissertation. The project supervisor will usually be a member of the Brunel Statistics or Financial Mathematics group. In some cases the project may be overseen by an external supervisor based in industry or another academic institution.

Read more about the structure of postgraduate degrees at Brunel and what you will learn on the course.

Read more about the structure of postgraduate degrees at Brunel and what you will learn on the course.