COMPM089 - Introduction to Deep Learning

This database contains the 2017-18 versions of syllabuses. Syllabuses from the 2016-17 session are available here.

Note: Whilst every effort is made to keep the syllabus and assessment records correct, the precise details must be checked with the lecturer(s).

CodeCOMPM089 (Also taught as COMPGI23)
Year4 (Masters)
PrerequisitesN/A
Term1
Taught ByNic Lane (100%)
AimsTo have a full understanding of the learning outcomes.
Learning Outcomes

At the conclusion of this module students should understand:

  1. The fundamental principles, theory and approaches for learning with deep neural networks
  2. The main variants of deep learning (such convolutional and recurrent architectures), and their typical applications
  3. The key concepts, issues and practices when training and modeling with deep architectures; as well as have hands-on experience in using deep learning frameworks for this purpose
  4. How to implement basic versions of some of the core deep network algorithms (such as backpropagation)
  5. How deep learning fits within the context of other ML approaches and what learning tasks it is considered to be suited and not well suited to perform

Content

This module will aim to teach students the fundamentals of modern multi-layered neural networks. It will cover the most common forms of model architectures and primarily the algorithms used to train them. The theory and principles will be presented, but this will go hand-in-hand with a focus on practical experience such as using existing frameworks and implementing (simplified versions) of core algorithms.

Students will be taught the basics of neural networks, convolutional networks, recurrent networks; and introduced to concepts such as: dropout, batch normalization, types of hyper-parameter optimization, distributed and constrained computing variants. Applications in the area of audio processing and image captioning and vision will be discussed.

Method of Instruction

Lectures

Assessment

The course has the following assessment components:

  • Coursework (100%) [comprises 3 assignments of 20%, 20% and 60%]

To pass this course, students must:

  • Obtain an overall pass mark of 50% for all sections combined.

Resources

Reading list available via the UCL Library catalogue.