Customer profile classification using transactional data

Authors: Apeh, E.T., Gabrys, B. and Schierz, A.C.

http://eprints.bournemouth.ac.uk/18459/

http://www.mirlabs.org/nabic11/

Pages: 37-43

Publisher: IEEE

ISBN: 978-1-4577-1122-0

DOI: 10.1109/NaBIC.2011.6089414

Customer profiles are by definition made up of factual and transactional data. It is often the case that due to reasons such as high cost of data acquisition and/or protection, only the transactional data are available for data mining operations. Transactional data, however, tend to be highly sparse and skewed due to a large proportion of customers engaging in very few transactions. This can result in a bias in the prediction accuracy of classifiers built using them towards the larger proportion of customers with fewer transactions. This paper investigates an approach for accurately and confidently grouping and classifying customers in bins on the basis of the number of their transactions. The experiments we conducted on a highly sparse and skewed real-world transactional data show that our proposed approach can be used to identify a critical point at which customer profiles can be more confidently distinguished.

This data was imported from DBLP:

Authors: Apeh, E.T., Gabrys, B. and Schierz, A.

http://eprints.bournemouth.ac.uk/18459/

http://ieeexplore.ieee.org/xpl/mostRecentIssue.jsp?punumber=6083215

Journal: NaBIC

Pages: 37-43

Publisher: IEEE

ISBN: 978-1-4577-1122-0

This data was imported from Scopus:

Authors: Apeh, E.T., Gabrys, B. and Schierz, A.

http://eprints.bournemouth.ac.uk/18459/

Journal: Proceedings of the 2011 3rd World Congress on Nature and Biologically Inspired Computing, NaBIC 2011

Pages: 37-43

ISBN: 9781457711237

DOI: 10.1109/NaBIC.2011.6089414

Customer profiles are by definition made up of factual and transactional data. It is often the case that due to reasons such as high cost of data acquisition and/or protection, only the transactional data are available for data mining operations. Transactional data, however, tend to be highly sparse and skewed due to a large proportion of customers engaging in very few transactions. This can result in a bias in the prediction accuracy of classifiers built using them towards the larger proportion of customers with fewer transactions. This paper investigates an approach for accurately and confidently grouping and classifying customers in bins on the basis of the number of their transactions. The experiments we conducted on a highly sparse and skewed real-world transactional data show that our proposed approach can be used to identify a critical point at which customer profiles can be more confidently distinguished. © 2011 IEEE.

The data on this page was last updated at 05:16 on February 19, 2020.