Analytics

Research in this area explores the study of data, methods for analysing data in different domains, and artificial intelligence techniques for improving our methods of working with large and heterogeneous data. This encompasses novel research on artificial intelligence, computer games, e-Sports, natural language processing and machine learning.

To find out more about research in this area, contact Professor Richard Wilson or read our most recent publications.

People

Dr James Cussens

Professor Peter Cowling

Professor Anders Drachen

Dr Dimitar Kazakov

Professor Dimitris Kolovos

Dr Daniel Kudenko

Dr Suresh Manandhar

Professor Richard Paige

Dr Christopher Power

Dr Siamak F. Shahandashti

Dr Tommy Yuan

Dr Peter Nightingale

 

Selected recent and ongoing projects

Crowd-Sourced Prediction of Plant Pest and Disease Occurrence

Contact: Dr Daniel Kudenko

Funders: InnovateUK, BBSRC

Partners: Growing Interactive

This project looked at developing innovative systems and visualisations that accurately predict pest and disease emergence on crops. The project statistically analysed the crowd-sourced pest and disease observations recorded in the Big Bug Hunt initiative in conjunction with meteorological information. Using meteorological information and weather forecasts allows growers to derive predictions for the current growing season that are specific to their location. The systems will form the basis of new services for customers of Growing Interactive, the horticultural industry and agriculture, and help growers to reduce losses due to plant pests and diseases.

 

 

MCTS for Business Decision Making

Contact: Dr Daniel Kudenko

Funders: InnovateUK

Partners: MooD International

Software company MooD International identified software techniques currently used in computer game operation that could potentially enhance the user experience of their business performance management solutions by ‘suggesting the next move’ in achieving business outcomes. Monte Carlo Tree Search (MCTS) is particularly promising but has been largely limited to the immediate domain of ‘game intelligence’, where it is highly successful.

The focus of the project is to apply and, where needed, adapt this technique to business decision making, which is a different and challenging domain. The ultimate goal is to enhance the functionality of MooD International’s business performance management solutions with a decision support engine based on MCTS.