Classifications and Applications in Big Databases
Tues December 2nd 2014, 1.00-4.30pm, RSS Errol Street, London.
1.00pm – Lunch
1.45pm – Mythbusting with latent class models, Dr. Matthew Sperrin
2.20pm – Ethnic differences in multiple health behaviours during pregnancy: An exploration using latent class analyses, Dr. Emily Petherick
2.55pm – Break
3.10pm – Who you know and who you are: using multiple-correspondence-analysis based clustering to examine social networks in the Great British Class Survey, Dr. Daniel Laurison
3.45pm – Modelling expert uncertainty in medical decision support, Prof. Jon Garibaldi
4.20pm – Final discussion/closing remarks
All are welcome. You must register in advance to attend. For registration details see: http://www.statslife.org.uk/events/eventdetail/303/14/classification-methods-and-applications-in-big-databases
More details about the speakers and their talks are given below.
See you there,
Pia Hardelid, Katie Harron and Annie Herbert
Speakers and abstracts:
“Modelling Expert Uncertainty in Medical Decision Support”, Prof. Jon Garibaldi (School of Computer Science, University of Nottingham)
Medical decision making is often difficult, requiring complex decisions in the precence of much uncertainty (both in data and in domain knowledge). In this talk, I shall present some of the recent work we have carried out in modelling uncertainty whilst performing various clustering and classification tasks to support medical decision making. In particular, I will focus on the variability exhibited by human decision makers, and how modelling this may lead to improved decision making.
“Who you know and who you are: using multiple-correspondence-analysis based clustering to examine social networks in the Great British Class Survey”, Dr. Daniel Laurison (Department of Sociology, London School of Economics)
Between 2011 and 2013, over 325,000 people went to the BBC website and took survey about class and identity. This created an unprecedentedly large set of entirely unrepresentative survey responses, which was paired with a small but reasonably representative survey with identical questions. Using these two sources, I examine the relationship between social position and social networks in Britain. I use Multiple Correspondence Analysis (MCA) on the smaller dataset, and generate clusters based on individuals’ positions in the primary plane of the MCA. I then match individuals from the large dataset to these clusters, and explore the ways the characteristics of the clusters differ across the two datasets.
“Ethnic differences in multiple health behaviours during pregnancy: An exploration using latent class analyses”, Dr. Emily Petherick (Born in Bradford Study, Bradford Institute for Health Research)
Pregnancy is a time of optimal motivation for many women to make positive behavioural change to improve their own health as well as the health of their unborn child. However, the majority of behaviour change models are applied in research and practice to single behaviours, for example quitting smoking or increasing physical activity. We present an example of the use of latent class analyses to further understand the grouping of these health behaviours and further examine the relationship between class membership and different social and demographic characteristics using data from the Born in Bradford cohort.
“Mythbusting with Latent Class Models”, Dr. Matthew Sperrin, (Centre for Health Informatics, University of Manchester)
Latent class models are often used to represent distinct subpopulations within a heterogeneous population. Ecological fallacy can arise when these latent classes are aggregated, leading to erroneous statements being made about a population. We present examples of this in the obesity epidemic, in longitudinal trajectories of disease risk factors, and in the ‘atopic march’ of allergy in childhood.
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|