果冻影院

Aims:

The overarching aim will be to introduce students to the most current research in machine learning. This will be of particular value to students planning to undertake PhD or Industrial Research.

Intended learning outcomes:

On successful completion of the module, a student will be able to:

1. Evaluate the differences between the various methods that achieve the goal of this particular topic and thus understand their strengths and weaknesses.
2. Implement an algorithm that achieves the aim of this particular topic.
3. Develop research skills that thus enable them听to听critically evaluate competing research papers that address the same problem (topic) and devise experiments as well as theoretical arguments enabling insightful comparison between methodologies.
4. Develop presentation and writing skills to analyse and explain modern machine learning methods.

Indicative content:

This module听is designed to introduce students to the most current research topics in machine learning. Such topics will correspond to 鈥渢rending鈥� topics within the last five years as represented in International Machine Learning Conferences.

The backbone of the module will be a series of lectures on a given set of selected topics for that year. As appropriate this may be supplemented by seminar-style class work where current research papers are read in common, discussed, and presented. There will be many potential 鈥渢rending鈥� topics that can be the subject for any given year, for example:

• Fairness and Artificial Intelligence
• Scalable Gaussian processes
• Modern variational inference
• Meta-Learning
• Bayesian Optimisation
• Multi-task Learning
• Mirror Descent
• Probabilistic Programming
• Inductive Logic Programming
• Machine Learning and Physical Sciences
• Privacy-preserving Machine Learning
• Submodular optimization

To name a few.

Requisites:

To be eligible to select this module as an optional or elective, a student must be registered on a programme and year of study for which it is a formally available.