Deep Learning Implementation and Optimisation module (CS51011)

Deep learning has revolutionised the way computers interpret information, powering applications from speech recognition and image analysis to recommendation systems and generative AI. This module introduces you to the practical use of deep learning frameworks, teaching you how to work with pre-trained models and adapt them to new challenges.

You’ll learn how modern neural networks operate, explore the architectures that underpin them, and gain hands-on experience with popular tools such as TensorFlow, Keras, and PyTorch. The emphasis is on applying existing models effectively — understanding how to fine-tune, evaluate, and optimise them to achieve accurate and efficient results in real-world contexts.

What you will learn

In this module, you will:

• study the structure and function of neural networks and deep learning architectures
• explore how pre-trained models can be used and adapted for specific tasks
• gain practical experience with frameworks such as TensorFlow, Keras, and PyTorch
• apply transfer learning techniques to real-world datasets
• evaluate and optimise deep learning models for performance and efficiency

By the end of this module, you will be able to:

• apply deep learning techniques to a range of applied problems
• assess and compare the performance of pre-trained models
• adapt existing models to new domains using transfer learning
• communicate technical results through structured reporting and evaluation

Assignments / assessment

• deep learning investigation project (100%)

Assessment is based on a practical investigation applying deep learning models to a real-world problem, supported by a written report.

This module does not have a final exam.

Teaching methods / timetable

You will learn through a combination of lectures, seminars, and lab-based workshops. Practical sessions focus on experimentation, giving you the opportunity to train, adapt, and evaluate neural networks using real or simulated datasets with academic guidance and peer collaboration.