Modeling charity donations using target selection for revenue maximization

J.M. Costa Sousa, da, S. Madeira, U. Kaymak

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review


This paper presents the results of one application of target selection in direct marketing: the mailing campaigns of a charity organization, where the clients are selected based on the expected amount of donation they are going to make. Target selection is an important data mining problem for which several modeling techniques have been used. Statistical regression, neural networks, decision trees, and clustering are the most utilized techniques. Fuzzy clustering can also be applied to target selection. In this paper, traditional and fuzzy techniques are compared by using cross-validation measures. The four techniques are applied based on recency, frequency and monetary value measures. The application to mailing campaigns of a charity organization, showed that fuzzy modeling obtains results similar to those of other classical target selection techniques.
Original languageEnglish
Title of host publicationThe 12th IEEE International Conference on Fuzzy Systems (FUZZ '03)
Place of PublicationPiscataway
PublisherInstitute of Electrical and Electronics Engineers
ISBN (Print)0-7803-7810-5
Publication statusPublished - 2003


Dive into the research topics of 'Modeling charity donations using target selection for revenue maximization'. Together they form a unique fingerprint.

Cite this