Undergraduate Programme and Module Handbook 2023-2024 (archived)
Module ENGI4607: Artificial Intelligence and Deep Learning
Department: Engineering
ENGI4607:
Artificial Intelligence and Deep Learning
Type |
Tied |
Level |
4 |
Credits |
10 |
Availability |
Available in 2023/24 |
Module Cap |
None. |
Location |
Durham
|
Tied to |
H100 |
Tied to |
H106 |
Tied to |
H108 |
Tied to |
H411 |
Tied to |
H412 |
Tied to |
H413 |
Tied to |
H911 |
Tied to |
H912 |
Tied to |
H913 |
Tied to |
H511 |
Tied to |
H512 |
Tied to |
H513 |
Tied to |
H711 |
Tied to |
H712 |
Tied to |
H713 |
Tied to |
H311 |
Tied to |
H312 |
Tied to |
H313 |
Tied to |
H811 |
Tied to |
H812 |
Tied to |
H813 |
Prerequisites
- MATH1551 Maths for Engineers and Scientists
Corequisites
Excluded Combination of Modules
Aims
- This module is designed solely for students studying Department of Engineering degree programmes.
- To provide underlying methodological and practical knowledge in the field of Artificial Intelligence (AI), machine learning and deep learning, covering a wide range of the modelling and computational techniques ubiquitous in real-world problems in companies, healthcare and bio-related applications, academia or the third sector.
- To enable students to approach complex ill-defined problems that require deep layers of learning, and understand how this relates to learning in nature.
- To equip students with the necessary knowledge and skills to work in the field of AI and to contribute to ongoing research in the area.
- To make students aware of best practices for fair and equitable use of AI.
Content
- Introduction of AI and Machine Learning
- Machine Learning Basics
- Classical Supervised Learning Algorithms
- Classical Clustering Algorithms
- Dimensionality Reduction (Feature Extraction) methods
- Deep Neural Networks
- Recurrent Neural Networks
- Autoencoder
- Generative Models
- Transfer Learning and fine tuning
- Review of some advanced methods
- Ethical and Bias Issues
Learning Outcomes
- The key principles of machine learning managing datasets and building models and core methodologies in relation to managing data and training models.
- An understanding of state-of-the-art deep neural network architectures and neural network architecture components.
- An understanding of the algorithms and approaches to design and evaluate deep neural networks.
- The key principles of ethical concerns and bias in AI for real-world applications, from an accountability perspective and as regards the application of algorithms (Ethics and Bias in AI).
- An ability to manage data and to select and apply appropriate algorithms to recognise patterns within the data, together with an ability to implement, analyse and compare learning algorithms.
- An ability to use modern deep learning libraries to design, train, validate and test deep neural networks.
- An ability to design appropriate neural network architectures suited for a given task or dataset.
- An ability to discuss implications of AI solutions in real-world applications and applying sensitivity analysis to given data (Ethics and Bias in AI).
- An ability to communicate technical information in the domain of artificial intelligence and machine learning .
- An ability to learn, understand, and visualise the underlying structure of datasets.
- An ability to design and implement state-of-the-art machine learning models.
- The scientific approach to design, training, validation, and testing of deep neural networks in a broad range of applications.
- An ability to appreciate positive and negative societal impact of AI.
Modes of Teaching, Learning and Assessment and how these contribute to
the learning outcomes of the module
- The module content is delivered in lectures and is reinforced by problem sheets, equipping students with the required problem-solving capability.
- Students are able to make use of staff 'Tutorial Hours' to discuss any aspect of the module with teaching staff on a one-to-one basis. These are sign up sessions available for up to one hour per week per lecture course.
- Formative and summative assignments encourage and guide independent study, and test the knowledge acquired and the students' ability to use this knowledge to solve problems.
Teaching Methods and Learning Hours
Activity |
Number |
Frequency |
Duration |
Total/Hours |
|
Lectures |
20 |
Typically 1 per week |
1 Hour |
20 |
|
Tutorial Hours |
As required |
Weekly sign-up sessions |
Up to 1 Hour |
10 |
|
Formative exercises and self-study |
|
|
|
70 |
|
Total |
|
|
|
|
|
Summative Assessment
Component: Coursework |
Component Weighting: 100% |
Element |
Length / duration |
Element Weighting |
Resit Opportunity |
Coursework |
|
100% |
|
■ Attendance at all activities marked with this symbol will be monitored. Students who fail to attend these activities, or to complete the summative or formative assessment specified above, will be subject to the procedures defined in the University's General Regulation V, and may be required to leave the University