Artificial Intelligence (with a Year in Industry) MSci (with Foundation Year)

The Foundation Year is a great opportunity if you have the ability and enthusiasm to study for a degree, but do not yet have the qualifications required to enter directly onto our degree programmes. A significant part of the Foundation Year focuses upon core skills such as academic writing at HE level, becoming an independent learner, structuring academic work, critical thinking, time management and note taking.

Successful completion of the Foundation Year will enable you to progress into the first year (Level C) of your chosen honours degree. Further details can be found here.

This is a broad introduction to the subject and you develop the theoretical knowledge, problem solving and practical skills that underpin AI:

• Introduction to Programming: You will learn the fundamentals of two programming languages.  Python is crucial to machine learning, and Java is a popular language for development and production.

• Introduction to AI: You start with classical approaches to AI, such as expert systems, as well as optimisation, the basics of finite state machines, and simple chatbots.

• Algorithm Design: Understanding algorithms and their creation is crucial to much of what you will do later in your degree and career, and you will learn this process from the ground up.

• Mathematics for AI: Linear algebra and calculus are used extensively in machine learning, and you will learn the basics of these during your first year.

• Taster of current and new developments: You will be taught by staff members with experience in diverse aspects of AI, giving you different perspectives on what work is being done in the field.

Your second year allows you to study specific topics in more depth, focusing on specific branches of theoretical and practically based AI subjects:

• Artificial Neural Networks: This is the most fundamental machine learning model, and you will learn how these are designed and trained.

• Software Development for AI: AI is not just about machine learning.  You will also learn in depth the principles of software development and being part of a team in industry.  Java and C++ are used as vehicles for this, as well as Unified Modeling Language (UML) for designing projects.

• Predictive Analytics: An important aspect of AI is making predictions, and you will learn more than one machine learning model, as well as how to properly pre-process data and build pipelines.

• Genetic Algorithms: Another important aspect of AI is metaheuristics, which is advanced optimisation to solve difficult problems, often ones that are NP-complete (or harder).  Genetic algorithms are the oldest metaheuristic, inspired by how creatures adapt to their environment.

• Natural Language Processing: There is an increasing demand in industry for candidates competent in NLP, which is the application of machine learning and information-theoretic techniques to human language. You will learn this in depth.

You spend your third year on a compulsory placement in industry.

Your fourth year helps you develop a deeper understanding of the specific aspects of AI and be able to critically select appropriate tools and techniques to solve specific problems:

• Machine Learning and AI at the edge: Edge ML involves a broad range of techniques to process data, often using deep learning, on smart devices.  You will gain access to such devices and use industry-level techniques and libraries.

• Embedded Systems, Physical Computing & Spatial Computing: Building on Edge ML, you will build machine learning into small devices, enabling the use of AI in everyday human interactions.

• Neuromorphic Computing: This is an emerging field of AI, using chips that mimic activities in the brain and nervous system, often using Very Large Scale Integration (VLSI) systems. 

• Robotics: Embedded systems and physical computing gives rise to robotics on a larger scale, and you will be able to implement machine learning into robotic systems.

• Computer Vision: AI is not just about making predictions on columnar data, but also about being able to work with high dimensional data, such as images, and you will learn how to apply deep learning on images (and by extension audio and other data), as well as the theory behind this.

• Human Machine Interaction: This ties into the knowledge you will gain of Edge ML and spatial computing, enabling you to understand the principles of how humans interact with various types of machines.

In your final year you will be studying with a great degree of autonomy. The Curriculum will be focused on applications of material that has been covered in prior years in a professional context:

• Advances in AI: The curriculum will be guided by the latest research in machine learning and individual staff knowledge and research.

• Advanced Machine Learning: You will learn the algorithms and mathematics behind many types of machine learning models, such as tree-based learning, support vector machines, spiking neural networks, deep Q-learning and convolutional neural networks.

• Internet Of Things: You will build upon your knowledge of embedded systems gained in the previous year, broadening it to the emerging internet-of-things, whereby physical things in the world are part of the internet.

• Intelligent Systems: Based on the latest advances in AI, you will learn various techniques in AI, not least using reinforcement learning to navigate complex environments.

• Finite State Machines: This side of the course will cover FSMs in great detail, such as deterministic vs non-deterministic finite automata, epsilon NFAs, converting between different types of models, and obtaining output from these models using Moore and Mealy machines.

• Game Theory: Although game theory started as a field in psychology, then formalised mathematically, it has more recently emerged as a field of AI.  It can indeed be viewed as a two- (or more) player alternative to reinforcement learning, obtaining optimal strategies for real-world scenarios, not least in industry and commerce.

• Bayesian Learning: At its heart, machine learning is implicitly about calculating probability distributions, but Bayesian Learning does this explicitly.  You will learn about Bayes Theorem and how this is used in Naive Bayes to perform predictive analytics, and this will be expanded to Bayesian Belief Networks for optimising complex systems.

• Group Project: Based on the wealth of knowledge throughout your degree, you will carry out a sizable student-led project on a topic of your choice.