This paper addresses focused information acquisition for predictive data mining. Asbusinesses strive to cater to the preferences of individual consumers, they often employpredictive models to customize marketing efforts. Building accurate models requiresinformation about consumer preferences that often is costly to acquire. Prior research hasintroduced many acirc; active learningacirc; policies for identifying information that is particularlyuseful for model induction, the goal being to reduce the acquisition cost necessary to inducea model with a given accuracy. However, predictive models often are used as part of adecision-making process, and costly improvements in model accuracy do not always result inbetter decisions. This paper develops a new approach for active information acquisition thattargets decision-making specifically. The method we introduce departs from the traditionalerror-reducing paradigm and places emphasis on acquisitions that are more likely to affectdecision-making. Empirical evaluations with direct marketing data demonstrate that for afixed information acquisition cost the method significantly improves the targeting decisions.The method is designed to be genericacirc; not based on a single model or inductionalgorithmacirc; and we show that it can be applied effectively to various predictive modelingtechniques
