DATA MINING
RSS Reading Local Group Meeting
4pm Thursday 24 February (Refreshments from 6pm)
Room 1L04
Department of Agriculture
University of Reading
Speakers:
DAVID ESP (National Grid Transco UK)
Data mining methods for demand forecasting at National Grid Transco
National Grid Transco UK owns and operates the country's main trunk
supply systems for both gas and electricity. This presentation considers
data-mining techniques for high-accuracy demand forecasting, based on a
generally superior type of neural network. The problem of optimal input
selection by genetic algorithm and other methods is addressed.
SIMON CUMMING (British Airways)
Data mining at British Airways
This presentation explores ways in which BA gains value from customer
data and investigates business complexity. Techniques including decision
trees, logistic regression, cluster analysis and self-organising maps
are used to measure customer value, segment customer data, predict
customer attrition and understand the uptake of new technologies such as
electronic tickets.
Directions are available at http://www.rdg.ac.uk/RSSlocal/
|