Unsupervised Optimal Fuzzy Clustering
IEEE Transactions on Pattern Analysis and Machine Intelligence
Selection of relevant features and examples in machine learning
Artificial Intelligence - Special issue on relevance
Automatic Construction of Decision Trees from Data: A Multi-Disciplinary Survey
Data Mining and Knowledge Discovery
Attribute Reduction of Rough Sets in Mining Market Value Functions
WI '03 Proceedings of the 2003 IEEE/WIC International Conference on Web Intelligence
Customer Targeting: A Neural Network Approach Guided by Genetic Algorithms
Management Science
Using Multivariate Statistics (5th Edition)
Using Multivariate Statistics (5th Edition)
Classification of proxy labeled examples for marketing segment generation
Proceedings of the 17th ACM SIGKDD international conference on Knowledge discovery and data mining
Hi-index | 0.00 |
Customer relationship management (CRM) aims to build relationswith the most profitable clients by performing customersegmentation and designing appropriate marketing tools. Inaddition, customer profitability accounting (CPA) recommendsevaluating the CRM program through the combination of partialmeasures in a global cost-benefit function. Several statisticaltechniques have been applied for market segmentations although theexistence of large data sets reduces their effectiveness. As analternative, decision trees are machine learning models that do notconsider a priori hypotheses, achieve a high performance, andgenerate logical rules clearly understood by managers. In thisarticle, a three-stage methodology is proposed that combinesmarketing feature selection, customer segmentation throughunivariate and oblique decision trees, and a new CPA function basedon marketing, data warehousing, and opportunity costs linked to theanalysis of different scenarios. This proposal is applied to alarge insurance marketing data set for alternative cost and priceconditions showing the superiority of univariate decision treesover statistical techniques.