Segmentation-based modeling for advanced targeted marketing

Abstract

+1 914 945 3000 apte, ebibeln, nramesh, pednault, fateh @us.ibm.corn ABSTRACTFingerhut Business Intelligence (BI) has a long and successful history of building statistical models to predict consumer behavior. The models constructed are typically segmentationbased models in which the target audience is split into subpopulations (i.e., customer segments) and individually tailored statistical models are then developed for each segment. Such models are commonly employed in the direct-mail industry; however, segmentation is often performed on an ad-hoc basis without directly considering how segmentation affects the accuracy of the resulting segment models. Fingerhut BI approached IBM Research with the problem of how to build segmentation-based models more effectively so as to maximize predictive accuracy. The IBM Advanced Targeted MarketingSingle Events TM (IBM ATM-SE TM) solution is the result of IBM Research and Fingerhut BI directing their efforts jointly towards solving this problem. This paper presents an evaluation of ATM-SE's modeling capabilities using data from Fingerhut's catalog mailings.

SOA-enabled compliance management: instrumenting, assessing, and analyzing service-based business processes

“…In addition to that, although the overall process is automatically orchestrated and activities have suitable IT support, in practice many tasks are still based on paper forms filled by doctors or nurses during their service and manually input only in a later stage. 1 …”

“…As for the root cause analysis, Grigori et al [11,12] focus on understanding, predicting, and preventing exceptions in business executions by using decision trees built upon workflow log files. In the same line of thought, Rozinat and van der Aalst [25] mine event logs for decision point analysis, Apte et al [1] focus on classification and prediction of customer behaviors, and Seol et al [31] use the inputs and outputs of each process to build decision trees for the analysis of the efficiency of processes. There are, however, no works that specifically address the problem of understanding compliance violations.…”

Tax Collections Optimization for New York State

“…Note that the equality corresponds to regression or estimation, whereas maximization corresponds to optimization, both over a large population. In our method, we do the estimation step using a scalable procedure for segmented linear regression; in each segment, R is estimated using regression in terms of action variables (Apte et al 2001, Natarajan andPednault 2002). Thus, the procedure discovers segments in the feature space, each of which is relatively uniform with respect to the effects of actions.…”

“…As its components do, it uses a scalable segmented regression engine (Apte et al 2001, Natarajan andPednault 2002) and a fast linear program solver (COIN-OR), both of which were developed primarily at IBM Research. These modules are incorporated in a unifying framework of C-RL; the solution provides a scalable system that is well suited for real-world deployments.…”

Implementation study: Using decision tree induction to discover profitable locations to sell pet insurance for a startup company

“…5 -11 Numerous studies have illustrated how data mining techniques can be specifi cally used to identify prospective customers. 1,2,5,[12][13][14] Decision tree algorithms, such as C4.5, 15 SBP 16,17 and others have also been used by many researchers to extract knowledge from databases that can be used by managers to make decisions. 18 In the case study in this paper, we make use of decision tree techniques for assisting a new pet insurance company to characterise its target market.…”