Contents addressed in this module may include, but not be limited to:

Introduction to Neural Networks:

Single-Layer Networks, linear discriminant functions, perceptron, XOR problem, Multi-Layer Perceptron, perceptron learning criteria, perceptron learning algorithm, and feedforward neural network architecture

Training Neural Networks (A selection of following topics)

Back propagation, Empirical risk minimization, regularisation, gradient descent, stochastic gradient descent, local and global extrema, Under-Fitting and Over-Fitting, early stopping, convergence rates for smooth convex optimisation problems, momentum acceleration methods, dropout, batch normalisation, modern optimization solvers, AdaGrad, RMSProp and Adam

Convolutional Neural Networks (CNNs)

Convolutional operations, convolutional neural networks, pooling, parameter sharing, and basic CNN architectures

CNN architectures

Convolutional neural network architectures, LeNets, AlexNets, VGGnets, Inception networks, and ResNets

Sequence modelling:

Recurrent Neural Networks (RNN): RNN architectures, exploding/vanishing gradients, and gated recurrent units

Long Short-Term Memory (LSTM) units, and bi-directional RNN

Encoder-decoder architecture, attention mechanism, transformer architecture, and neural machine translation

Pre-trained Networks and Transfer Learning

Application of Deep Learning

Practical Assessment – A practical assessment that provides hands on experience in using AI-purposed libraries and tools for coding of algorithms. Students will research, analyse, design, implement, test and document an ANN model solution to a case study (Learning Outcomes 1, 2 and 3).

Written Report – Students will be expected to analyse a given task scenario (e.g. Digital Health monitoring, Crowd and traffic Management, or Smart Homes) in order to define a business problem they need to design for, develop for the practical assessment (utilising an ANN model using the theories and practices covered in the module). This work will be assessed by a written report (Learning Outcomes 1, 2 and 3).

Theory will be delivered via lectures and supported by practical classes, seminars and discussion groups. In addition, you will be provided with a range of resources for independent study such as case studies, academic papers and industry stories. There will be a mixture of practical and theoretical formative exercises which will help you build your knowledge and confidence as well as preparing you for the summative assessment.

1. Demonstrate a critical understanding of the benefits and limitations of an Artificial Neural Network (ANN), architecture, techniques and Algorithms in comparison to other machine learning methods

Knowledge and Understanding, Learning

2. Develop and apply ANN models to specific technical and scientific problems

Analysis, Reflection, Problem Solving

3. Critically appraise the application of Deep Learning with data and tasks from a specific domain (e.g. health, or finance)

Problem Solving

JetBrains PyCharm (IDE for Python) or Visual Studio Code or similar IDE

Python, miniconda, Pytorch

TensorFlow,Keras

Datasets from Kaggle, Weka, MIMIC or similar sources

These are indicative only. Texts are updated on an annual basis and when you start to study this module, you will be referred to an online reading list, currently provided through Keylinks. You are advised not to buy any textbooks for this module without checking the online reading list.



Theobald, A. (2021), Machine Learning for Absolute Beginners: A Plain English Introduction (Third Edition): 1 (Machine Learning with Python for Beginners), Independently published

Geron, A. (2022), Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow 3e: Concepts, Tools, and Techniques to Build Intelligent Systems, O'Reilly Media; 3rd edition

Huyen, C. (2022), Designing Machine Learning Systems: An Iterative Process for Production-Ready Applications, O'Reilly Media; 1st edition

Tombs, N. (2022), Reflections of a Robot: Dialogues on Artificial Intelligence, Independently Published

Clarke, E. (2022), Everything Data Analytics-A Beginner's Guide to Data Literacy: Understanding the Processes That Turn Data Into Insights (All Things Data), ¿Kenneth Michael Fornari

Aspen-Taylor, S. (2022), Data and Analytics Strategy for Business: Unlock Data Assets and Increase Innovation with a Results-Driven Data Strategy, Kogan Page; 1st edition

Theobald, A. (2022), Data Analytics for Absolute Beginners: A Deconstructed Guide to Data Literacy: (Introduction to Data, Data Visualization, Business Intelligence & ... Science, Python & Statistics for Beginners), Independently published¿¿

Raschka, S. (2017) Python machine learning: machine learning and deep learning with Python, scikit-learn, and TensorFlow.

Raschka. S. (2016) Python Machine Learning. Packt 3

Lewis, N. D. (2016) Deep Learning Step by Step with Python 4

Rashid T. (2016) Pattern Recognition and Machine Learning, Springer 6

Kowalski. R. et. al. (2011) Data mining: practical machine learning tools and techniques.¿3rd edn. London; Amsterdam: Morgan Kaufmann

This module establishes a fundamental understanding of Artificial Neural Networks (ANN) characteristics, techniques, and algorithms that are inherent in the solution of complex, real-world problems. As a student on this module, you will learn to design and implement ANN for real-world applications using professional ML tools. The assessment is specifically focused on exploration of algorithms in the real world and will use a case study for which you must investigate and find solutions for.