¹û¶³Ó°Ôº

Aims:

This module provides a broad introduction to probabilistic modelling. The module starts with a review of probability theory and Bayesian reasoning, before moving to more advanced techniques for approximate Bayesian inference (variational inference, expectation propagation, sampling, to name a few) and finally covering Bayesian machine learning models such as Gaussian processes. These core principles and algorithms will be presented alongside example applications.Ìý

The aims are to:Ìý

• Provide an understanding of fundamentals of probabilistic models and their applications.ÌýÌý
• Capacitate Machine Learning and Artificial Intelligence practitioners in the development and deployment of uncertainty quantification and management in machine learning pipelines.ÌýÌý
• Capacitate those individuals to be effective team players in interdisciplinary research groups/ organisations and institutions that utilise such modelling approaches.

Intended learning outcomes:

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

1. Demonstrate understanding of the fundamental basic principles of probability theory and Bayesian inference.Ìý

2. Implement and use key concepts, issues, and practices when training and modelling with probabilistic models.Ìý

3. Solve data challenges spanning different application domains and core learning tasks with Bayesian machine learning algorithms.Ìý

4. Demonstrate their acquired skills by attacking several real-world challenges using the techniques learned.Ìý

Indicative content:

The following are indicative of the topics the module will typically cover:

• Introduce to the fundamentals of probability theory and Bayesian modelling learning alongside their applications in the real-world.
• Probabilistic Modeling (e.g., graphical models, latent variable models, hidden Markov models.)Ìý
• Frequentist Inference.Ìý
• Bayesian Inference (e.g., variational inference, expectation propagation, sampling.)Ìý
• Bayesian Machine Learning (e.g., Bayesian linear regression, Gaussian processes.)Ìý

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 formally available.