This module offers a comprehensive exploration of foundational Bayesian algorithms within the realm of probabilistic machine learning, equipping you with cutting-edge techniques applicable across diverse domains including natural language processing, image recognition, and fraud detection. You'll delve into fundamental Bayesian Inference concepts, including prior and posterior distributions, Bayesian estimation, Bayes factor, model selection, and forecasting.
You’ll gain knowledge of various posterior sampling algorithms and see how to apply them through real-world instances in linear regression and classification. You’ll also learn about the latest trends in the field including variational Bayes and online learning. Through a combination of lectures and practical computer-based sessions, you’ll gain hands-on experience and theoretical insights, and gain a deep understanding of probabilistic machine learning methodologies.
Lecture 32, Computing classes 16
Problem sheets worth 40%.
Report including code worth 60%.
Reassessment Method: like-for-like
Including composite form of reassessment for failed portfolio component – written single problem sheet
On successfully completing the module, students will be able to:
Critically appraise the well-established principles and methods of Bayesian machine learning, and of the way those principles are built up and connected.
Deploy established approaches accurately to solve problems using skills in the following areas: derivation of posterior distributions
computation of posterior summaries, including the predictive distribution
construction of Bayesian hierarchical models and their estimation using Markov chain Monte Carlo methods
online regression
online classification
online machine learning
critical evaluation and interpretation of software output.
Formulate and identify the limits of the theory, and how this influences analyses and interpretations based on that knowledge.
An ability to use a range of established Bayesian machine learning techniques to initiate and undertake critical analysis of information, and to design solutions to problems arising from that analysis.
Have an ability to effectively communicate information, arguments, and analysis, in a variety of forms, to specialist and non-specialist audiences.
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.