Level 3 Diploma in Data Science

This unit provides learners with an introduction to Python programming for data science. The unit assumes no prior knowledge of coding or of Python and so starts by explaining the basics of Python, its design philosophy, syntax, naming conventions and coding standards.

The unit then introduces the basic Python data types of integers, floats, strings, complex numbers and booleans and explains how these data types can be created, changed, manipulated, and calculated using standard mathematical functions, logical operators, and Python’s built-in methods and functions. The unit also introduces more complex data structures critical to many data analytics and data science tasks, such as “lists”, “tuples”, “sets”, and “dictionaries”.

The unit explains how to use control and flow statements such as branching and looping as well as the basics of writing user-defined Python functions – all the ingredients needed to later perform data analysis and to code data science models successfully.

This unit introduces the learner to basic charts and visualisations and how to create and interpret them. The unit starts by explaining why visualisations are critical when understanding data and what makes a good and a poor visualisation.

The unit introduces learners to a number of basic chart and plot types, explaining their purpose, how to interpret them and explains when they should and should not be used. The unit then focuses on the technology used to produce charts and visualisations in Python, using Seaborn, Matplotlib and other Python libraries.

With modern software, packages, and programming languages, it is too easy for aspiring data scientists to rely on these tools to calculate descriptive statistics for them. It is critical for the modern data scientist to not only be able to interpret descriptive statistics, but also understand them and know how they are calculated. A lack of knowledge and the inability to interpret statistics correctly often leads to erroneous decisions being made which can have serious negative consequences.

This unit aims to provide learners with an introduction to descriptive statistics and methods which are key for data analysis and data science. This unit introduces different types of data and descriptive statistics from measures of centre, various measures of spread (including range, percentiles, variance and standard deviation), measures of symmetry (skewness and kurtosis) and measures of joint variability (correlation and covariance). The unit also explains which descriptive statistics can be calculated for the data measured on different scales. In this unit, learners will gain first-hand experience and practice of calculating descriptive statistics for small data sets manually.

This unit serves as the introduction to the core concepts of data analytics.

The unit will help learners to differentiate between the roles of a Data Analyst, Data Scientist and Data Engineer. Learners will also be able to summarize the data ecosystem such as databases and data warehouses and learn about major vendors within the data ecosystem and explore the various tools.

The unit also introduces learners to the fundamental tasks and processes in the data discovery process such as data cleaning, methods for dealing with data quality and methods for standardising data ready for analysis.

This unit introduces basic data analysis with Python. Learners are introduced to core concepts such as Pandas DataFrames and Series, merging and joining data.

This unit also builds on previous units by teaching how to import data, using Python to create descriptive statistics for analysis and interpretation. The unit also teaches learners how to use Python when preparing data for machine learning models by improving data quality and standardising data.

This unit provides a high-level overview (rather than a deep dive) of the three main types of machine learning: supervised learning, unsupervised learning, and reinforcement learning. The unit discusses the use-cases and real-world problems the various methods can be applied to, summarises the key-features of the different methods, as well as the challenges of each method.

This unit introduces the many steps and processes involved when building and evaluating machine learning models. The unit explains the core elements of the machine learning process from how to prepare data to selecting the correct machine learning algorithm to the importance of splitting data into training, test, and validation datasets to avoid the pitfalls of under and overfitting. The unit also covers how to identify and correct class imbalance and discusses when such approaches are needed. Many of the machine learning models that are encountered are supervised classification models and so the unit introduces the common performance metrics as well as how to interpret them. Finally, the unit discusses briefly how to deal with model bias and variance.

This unit introduces the basic theory of simple linear regression models that are critical to the ability to predict the value of one continuous variable based on the value of another. Learners will be able to estimate the line of best-fit by calculating the regression parameters and understand the accuracy of the line of best-fit. The unit also introduces extensions to simple linear regression by introducing multiple and polynomial regression models to examine relationships between multiple variables. The unit explains how to build simple, multiple, and polynomial linear regression models using Python and libraries such as scikit-learn.

This unit introduces logistic regression and its application as a classification algorithm. The unit explores the basics of binary logistic regression via the logistic function, the Odds ratio, and the Logit function. The unit also explains the differences between linear and logistic regression. Learners will learn how to build and visualise a logistic regression model using Python. The unit will teach learners when it is relevant to choose logistic regression over linear regression, how to interpret the results of logistic regression correctly and how to choose the best logistic model that describes the relationship under question.

This unit introduces the basic theory and application of decision trees. The unit explains how basic classification trees using the standard ID3 decision-tree construction algorithm are built and how nodes are split based on information theory concepts such has Entropy and Information Gain. The learner will also build and evaluate decision tree models in Python.

This unit introduces an unsupervised machine learning algorithm: k-means clustering. The unit aims to provide learners with the intuition behind k-means clustering algorithm and how to find the optimal number of clusters. Finally, the learner will also build and evaluate k-means methods in Python and will learn how visualise the clusters.

This unit aims to provide learners with an introduction into an emerging area of data science – synthetic data and its application to data privacy and security. Data collected by companies (such as Google, Facebook, Twitter) as well as governments, are a key resource in today’s information age. However, the leaking and inadvertent disclosure of data poses a serious threat to individual privacy. The unit introduces data privacy, the need for privacy and the legislative landscape. The unit explores traditional means of providing data privacy from anonymisation and encryption, before introducing the learner to the concept of differential privacy and the fundamental challenges of balancing data privacy with data utility.

This unit aims to provide learners with an introduction into another emerging area of data science – graphs and graph data science. This unit provides a gentle introduction to the field of graph theory which underpins all modern graph databases and graph analytics. The unit also covers the graph ecosystem, introducing Knowledge Graphs, Labelled Property Graphs and RDF graphs for data storages and processing. The unit introduces graph algorithms which are used to model, store, retrieve and analyse graph-structured data.

This unit introduces learners to the field of data science from the birth of artificial intelligence and machine learning in the late 1950s to the dawn of the “big data” era in the early 2000s, to the current applications of AI, machine learning and deep learning and the various challenges associated with them