eprintid: 107
rev_number: 4
eprint_status: archive
userid: 5
dir: disk0/00/00/01/07
datestamp: 2007-06-19
lastmod: 2015-05-29 19:47:27
status_changed: 2009-04-08 16:54:15
type: report
metadata_visibility: show
item_issues_count: 0
creators_name: Challet, Damien
creators_name: Krause, Andreas
contributors_name: Adjali, Iqbal
contributors_name: Wilson, Eddie
title: Validation of agent-based models
ispublished: pub
subjects: retail
studygroups: esgi56
companyname: Unilever
full_text_status: public
abstract: The automatic collection of customer transaction data, through either online shops or reward cards, is producing very large databases which contain much information about consumer behaviour. What kind of information and how exploitable it is are very relevant questions. Two approaches are being used. Either one concentrates on individual behaviour and tries to apply various theoretical frameworks and results of the literature on discrete choice, or one uses clustering algorithms in order to determine several classes of customers. The very existence of such categories is likely to be the result of social interactions and influences. The literature on discrete choice cannot easily be generalised to networked interactions, which are known to be widely present in various contexts. Another approach is to use toy models of individual behaviour and concentrate on global, aggregate quantities such as market share or demand fluctuations. This raises the question of how to validate such kind of model, hence the request of Unilever. The latter should also be understood with respect to the contribution of ESGI 2004, where a very sophisticated agent-based model of consumer behaviour was proposed (but not much studied).
problem_statement: There are two types of influences on consumer behaviour: global, through advertising and price setting, and local, via a network of social imitation. Unilever was most interested in knowing how to detect the presence of social imitation networks in consumer behaviour and how to extend and validate agent-based models to include social imitation. We proposed to apply recent results and managed to detect the presence of a social imitation network in a dataset of soap-buying by Swiss customers, thanks to a very simplified agent-based model.
date: 2007-04-17
date_type: published
pages: 9
citation:   Challet, Damien and Krause, Andreas  (2007) Validation of agent-based models.  [Study Group Report]     
document_url: http://miis.maths.ox.ac.uk/miis/107/1/Unilever-ABM-Report.pdf