Research projects in the Artificial Intelligence Group
The group undertakes a wide number of projects, often in collaboration with other groups at UWE Bristol. A list of our previous projects is also available. Recent or current projects include:
Intelligent Statistical Disclosure Control
A long term collaboration with the UK Office of National Statistics applying meta-heuristics to the problem of how to protect confidentiality in publicly available datasets (e.g. employment statistics, industrial outputs) while maximising the usefulness of what is published. Led by Jim Smith and Alistair Clark, with Martin Serpell.
Hypergraph Partitioning with Memetic Algorithms
HEFCE QR Funded: 2017
Investigating novel methods for improving the way that huge problems can be sub-divided to make them more tractable for optimisation. RA: Richard Preen. Supervisor: Jim Smith.
Design Mining: A Microbial Fuel Cell Pilot Study
Use of machine learning and 3D printing within the engineering design process to exploit novel materials/technology and enhance creativity under an agile-like system, with Ioannis Ieropoulos and John Greenman. RA: Jiseon You, Richard Preen. Supervisor: Larry Bull.
Creating Machine Intelligence with Intelligent Interactive Visualisation
UWE / Montvieux Ltd.: 2016-2019
This collaborative PhD project aims to resolve three of the challenges to building a predictive event analysis systems. Firstly, the encoding of event data in a format that is suitable for neural network input. Secondly, the training of neural networks for forecasting events. Finally, the use of introspective techniques to identify learning patterns and explain predictions. PhD Student: Sinclear-Emmanuel Smith. Supervised by Jim Smith and Phil Legg.
Enhanced Personal Situational Awareness (ePSA)
Vice Chancellor's Early Career Researcher Award: 2015-2016
Investigation into how machine learning and visual analytics techniques can be better utilised to assist non-expert users in understanding information sharing for networked devices. PI: Phil Legg.
Improved Search Methods for Improving Query Relevance and Accuracy
KTP with Paxport / Multicom: 2016-2017
This project developed novel combinations of Machine Learning in Recommender Systems tools to improve the relevance of the results returned by the company's' Find-and-Book service, which is used by travel agents worldwide. KTP Associate: Pedro Ferreira. Supervised by Chris Simons and Jim Smith.
Embodied Evolutionary Computing Design: Vertical Axis Wind Turbine Case Study
Leverhulme Trust: 2014-2015
Use of evolutionary computing to design hard to formulate/simulate physical systems through 3D printing.RA: Richard Preen. Supervisor: Larry Bull.