Deep learning is an approach to machine learning developed over the past several decades that draws heavily on our knowledge of the human brain, statistics and applied mathematics. In recent years, deep learning has grown tremendously in popularity and usefulness, largely due to more powerful computers, larger datasets and techniques to train deeper networks.
You will investigate modern deep neural networks starting from the fundamentals of Artificial Neural Networks (ANNs). You will start by learning what is an artificial neuron and architectures like multilayer perceptron, expanding on algorithms for training ANNs such as gradient descent and backpropagation.
You will examine applications of deep learning with particular attention to coding aspects during lectures and practical activities. You will learn to use the most popular techniques to evaluate the performance of a model on real data. For this, you will be introduced to different types of deep neural networks, from convolutional networks to recursive neural networks, and autoencoders to address different problems with a particular focus on computer vision topics.
Lectures, classes, seminars: 38
Spring (September Start)
Autumn (January Start)
Individual Mini project and 5 minutes individual presentation worth 60%
Computer-based invigilated mini-practical worth 40%
Reassessment - Single instrument
100% Practical assessment (mini-practical project and video presentation)
On successfully completing the module, students will be able to:
Design Deep Learning pipelines with best practices.
Build applications with the deep learning architecture with real-world data.
Compose Deep Learning models from state-of-the-art software packages with real data.
Critically analyse deep learning algorithms and their implementations based on a systemic understanding of the trade offs.
University of Kent makes every effort to ensure that module information is accurate for the relevant academic session and to provide educational services as described. However, courses, services and other matters may be subject to change. Please read our full disclaimer.