Most industry analysis starts with exploratory data analysis and a thorough study of this will help learners to perform data health checks and provide initial business insights.

The module will help the learner to understand and perform descriptive statistics and present the data using appropriate graphs/diagrams and serves as a foundation for advanced analytics.

This module also introduces the basics of programming in R and Python, the most commonly used languages used for data science.

The module culminates in practices related to data management, which is essential for both exploratory data analysis and advanced analytics. In particular, the module focuses on SQL as a highly practical language for data preprocessing, and addresses ways to connect SQL with R and Python tools, as well as learning the skills required to prepare data for machine learning and efficient data modelling.

Core Reading List:

R for Data Science: Import, Tidy, Transform, Visualise, and Model DataPaperback – 25 July 2016

by Garrett Grolemund (Author), Hadley Wickham (Author)

Hands-On Exploratory Data Analysis with Python: Perform EDA techniques to understand, summarise, and investigate your data Paperback – 27 Mar. 2020

by Suresh Kumar Mukhiya (Author), Usman Ahmed (Author)

Supplementary Reading List:

Exploratory Data Analysis with R

Radhika Datar, Harish Garg

Publisher: Packt Publishing (31 May 2019)

ISBN: 178980437X

This module provides learners with an in-depth understanding of the statistical distribution and hypothesis testing in a practical approach to getting things done.

Statistical distributions include Binomial, Poisson, Normal, Log-Normal, Exponential, t, F, and Chi-Square. Parametric and non-parametric tests used in research problems are covered in this unit.

The module will help learners to formulate research hypotheses, select appropriate tests of hypotheses, write primarily R programs to perform hypothesis testing, and to draw inferences using the output generated. After taking this course, students will understand the broad directions of statistical inference and use this information for making informed choices in analysing data.

Core Reading List:

Statistical Inference For Everyone

Copyright Year: 2017

Brian Blais, Bryant University

This module provides a strong foundation for predictive modelling. Its

objective is to define the entire modelling process with the help of real life

case studies.

Many concepts in predictive modelling methods are common and, therefore,

these concepts will be covered in detail in this module.

Students will learn how to carry out exploratory data analysis to gain

insights and prepare data for predictive modelling, an essential skill valued

in many industries.

The module also builds on information covered in the module Exploratory

Data Analysis to include hands-on applications of the summarization and

visualisation of datasets through plots to present results in compelling and

meaningful ways.

Core Reading List:

Mastering Predictive Analytics with R - Second Edition James D. Miller, Rui Miguel Forte Publisher Packt Publication date: August 2017

Predictive Analytics with Python, 1st Edition Alvaro Fuentes Publisher Packt

PowerBI and Excel are fundamental parts of the data analytics toolkit. A strong understanding of these also provides a basis for more advanced data analytics with other techniques and technologies. In this unit, learners will gain experience in collecting, processing, analysing, and communicating with data using Excel. In addition, data visualisation is a powerful way to communicate meaning in data and support business decision-making. This unit will cover the main commercial tools used in data visualisation such as Tableau and Power BI, enabling learners to create a wide range of graphs, charts, and dashboards and use them appropriately in context.

This unit provides learners with an opportunity to apply key knowledge and skills through project work. They will be able to select a project from a specific domain and will be required to carry out various data management, exploratory data analysis, data visualisation and predictive modelling tasks. The Data Science in Practice work should deepen their engagement with this material, and should prepare students for engaging fully with contemporary research methods in data science.

Reading List

(General):

• Gao, G., Mishra, B., & Ramazzotti, D. (2018). Causal data science for financial stress testing. J. Comput. Sci., 26, 294-304.

• Chen, H., Lundberg, S.M., & Lee, S. (2018). Hybrid Gradient Boosting Trees and Neural Networks for Forecasting Operating Room Data. ArXiv, abs/1801.07384. (Machine Learning)

• Miller, James D. and Rui Miguel Forte. Mastering Predictive Analytics with R: Machine Learning Techniques For Advanced Models. Second Ed. Birmingham: Packt, 2017.

• Fuentes, Alvaro. Mastering Predictive Analytics with Python. Birmingham: Packt, 2018. (Data Analytics in Business)

• Hands-On Exploratory Data Analysis with R: Become an Expert in Exploratory Data Analysis Using R Packages, Radhika Datar and Harish Garg, 1st Edition. (Packt Publishing, 2019). 266 pages

• Hands-On Exploratory Data Analysis with Python: Perform EDA Techniques to Understand, Summarize, and Investigate Your Data, Suresh Kumar Mukhiya and Usman Ahmed, 1st Edition. (Packt Publishing, 2020). 352 pages.

This module builds on the concepts introduced in the module Fundamentals of Predictive Modelling.

In this module, learners are introduced to model development for categorical dependent variables. Binary dependent variables are encountered in many domains such as risk management, marketing and clinical research and this unit covers detailed model building processes for binary dependent variables. Additionally, a primary goal of the module is for students to be able to select and successfully apply appropriate advanced regression models in applied settings.

The module will culminate with multinomial models and ordinal scaled variables.

Core Reading List:

Mastering Predictive Analytics with R - Second Edition James D. Miller, Rui Miguel Forte Publisher Packt Publication date: August 2017

Predictive Analytics with Python, 1st Edition Alvaro Fuentes Publisher Packt

Data reduction is a key process in business analytics projects. In this module, learners will learn data reduction methods such as Principal Component Analysis, Factor Analysis and Multidimensional Scaling.

Students will develop skills related to the formation of segments using cluster analysis methods. Additionally, students will analyse segments, the process of which is a key technique for large groups of data as intrinsic information appears in detail once segmented thoughtfully.

Required Reading Material:

• Applied Unsupervised Learning with R Publisher: Packt R Copyright Year: March 2019 ISBN 9781789956399 Alok Malik, Bradford Tuckfield

In this module, time series forecasting methods are introduced and explored. Students will gain a working knowledge of the nature and processes used in relation to time series data and confidently recognize and understand trends that exist within that data. This information will be used to make predictions or forecasts.

Students will analyse and forecast macroeconomic variables such as GDP and inflation. Additionally, students will work with complex financial models using ARCH and GARCH, ARIMA, time series regression, exponential smoothing, and other models.

Core Reading List:

Hands on Time Series Analysis with R

Rami Krispin

Publisher: Packt

Copyright Year: May 2019

Machine learning algorithms are new generation algorithms used in conjunction with classical predictive modelling methods.

In this Machine Learning 1 module, learners will understand applications of the Support vector machine, K Nearest Neighbours and Naive Bayes algorithms for classification and regression problems. Additionally, students will develop practical machine learning and data science skills including theoretical basics of a broad range of machine learning concepts and methods with practical applications to sample datasets.

Reading List:

Introduction to Machine Learning with Python: A guide for Data Scientists, Andreas Müller and Sarah Guido, 1st Edition. (O’Reilly Media, 2016).

Machine learning algorithms are new generation algorithms used in

conjunction with classical predictive modelling methods.

Machine learning algorithms are new generation algorithms used in

conjunction with classical predictive modelling methods. In this Machine Learning 2 module, students build on the knowledge gained from Machine Learning 1 and will go on to understand applications of decision trees and random forest algorithms, and neural networks for classification and regression problems. Additionally, students will develop practical machine learning and data science skills including theoretical basics of a broad range of machine learning concepts and methods with practical applications to sample datasets.

Reading List:

Introduction to Machine Learning with Python: A guide for Data Scientists, Andreas Müller and Sarah Guido, 1st Edition. (O’Reilly Media, 2016).

In this module, students will look at analysing unstructured data such as that found on social media, newspaper articles, videos, and more.

Specifically, students will look at text techniques for text mining and natural language processing using R and Python code to produce graphical representations of unstructured data and carry out sentiment analysis.

This module focuses on learning key concepts, tools, and methodologies for natural language processing and emphasises hands-on learning through guided tutorials and real-world examples.

Core Reading List:

Text Mining with R

Julia Silge and David Robinson.

O’Reilly

Natural Language Processing with Python

Steven Bird, Ewan Klein and Edward Loper.

O’Reilly

The Applied Data Science Practicum requires learners to investigate a real-world problem in the last phase of the MSc Data Science course. Its objective is to help students appropriately apply the concepts, techniques and tools learned from the Postgraduate Certificate and Diploma parts of the course to a real-world scenario.

Students typically choose a problem from a particular business or social domain after discussing it with the course instructor(s). They have the option of working on a real-world problem from their own organisation and working with a mentor in conjunction with their course supervisor. All external expert supervisors and projects need to be approved by the instructor(s) to ensure that the analytic question is appropriately scoped and technically challenging and that the solutions are rigorous and of high quality.

Students are required to solve an analytically complex research problem. Once the problem has been approved by the instructor(s), the student conducts a literature review of prior work in the field. Then, they conduct an exploratory data analysis, hypothesis testing, and research design and use a range of classical and/or modern machine learning modelling methods to predict outcomes and provide actionable insights and recommendations. Depending on the problem, the students may build dashboards or other artifacts as part of this work.

A key part of the project is to communicate the output of the learner’s research to technical and non-technical audiences through written, verbal and visual means.