Wyniki wyszukiwania - BazTech

1

Uplift Modeling in Direct Marketing

Rzepakowski P., Jaroszewicz S.

Journal of Telecommunications and Information Technology

|

2012

|

nr 2

43-50

EN

Marketing campaigns directed to randomly selected customers often generate huge costs and a weak response. Moreover, such campaigns tend to unnecessarily annoy customers and make them less likely to answer to future communications. Precise targeting of marketing actions can potentially results in a greater return on investment. Usually, response models are used to select good targets. They aim at achieving high prediction accuracy for the probability of purchase based on a sample of customers, to whom a pilot campaign has been sent. However, to separate the impact of the action from other stimuli and spontaneous purchases we should model not the response probabilities themselves, but instead, the change in those probabilities caused by the action. The problem of predicting this change is known as uplift modeling, differential response analysis, or true lift modeling. In this work, tree-based classifiers designed for uplift modeling are applied to real marketing data and compared with traditional response models, and other uplift modeling techniques described in literature. The experiments show that the proposed approaches outperform existing uplift modeling algorithms and demonstrate significant advantages of uplift modeling over traditional, response based targeting.

2

Mining interesting rules and patterns for salt sensitivity of blood pressure

Korzeń M., Klęsk P., Ciechanowicz A., Kaczmarczyk M., Widecka K., Jaroszewicz S.

Metody Informatyki Stosowanej

|

2008

|

nr 3 (Tom 16)

149--157

EN

Given a medical data set containing genetic description of sodium-sensitive and nonsensitive patients, we examine it using several techniques: induction of decision rules, naive Bayes classifier, voting perceptron classifier, decision trees, SVM classifier. We specifically focus on induction of decision rules and so called Pareto-optimal rules, which are of large interpretative value for physicians. We find statistically relevant combinations of attributes, which affect the sodium sensitivity.

3

Cross-selling models for telecommunication services

Jaroszewicz S.

Journal of Telecommunications and Information Technology

|

2008

|

nr 3

52-59

EN

Cross-selling is a strategy of selling new products to a customer who has made other purchases earlier. Except for the obvious profit from extra products sold, it also increases the dependence of the customer on the vendor and therefore reduces churn. This is especially important in the area of telecommunications, characterized by high volatility and low customer loyalty. The paper presents two cross-selling approaches: one based on classifiers and another one based on Bayesian networks constructed based on interesting association rules. Effectiveness of the methods is validated on synthetic test data.