Abstract
Compared to the classical scoring approach, the difficulty with net scoring is that the target variable, i.e. the uplift, is not defined for an individual observation. Rather, the impact of a treatment is measured by a comparison of structurally identical groups of observations which have (target group) or have not (control group) received the treatment. The underlying problem is that an observation cannot be treated and not treated at the same time. Due to this interaction of the response and the treatment variable, gross scoring methods are not directly applicable, yet they present the basis from which to move on. In this chapter, several statistical methods for net scoring are presented. Firstly, a general and formal description of the net scoring problem is provided. Then, a wide variety of statistical methods for uplift modeling are presented and their respective advantages and disadvantages are described. The two final sections deal with appropriate methods for responses or treatments that are not binary, contrary to what is assumed for most parts of the book.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
The notion uplift-uplift comes to mind: It nicely emphasizes the second order nature of net modeling and sticks in mind.
- 2.
The trade-off between stability and significance depends on the problem at hand.
References
F. Alemi, H. Erdman, I. Griva, and C.H. Evans. Improved statistical methods are needed to advance personalized medicine. Open Transl. Med. Journal, 1:16–20, 2009. https://doi.org/10.2174/1876399500901010016.
P. Austin. Using ensemble-based methods for directly estimating causal effects: An investigation of tree-based g-computation. Multivariate Behavioral Research, 47:115–135, 2012.
I. Bose and X. Chen. Quantitative models for direct marketing: A review from systems perspective. European Journal of Operational Research, 195(1):1–16, 2009.
P. Bremaud. An Introduction to Probabilistic Modeling. Springer, 2012.
M.D. Chickering and D. Heckerman. A decision theoretic approach to targeted advertising. 2000. in Boutilier, C., and Goldszmidt, M. (Eds.), Uncertainty in Artificial Intelligence, Proceedings of the Sixteenth Conference, Morgan Kaufman, San Mateo, California.
H. Cramér. Mathematical Methods of Statistics. University Press, Princeton, 1945.
M. Falk, H. Fischer, J. Hain, F. Marohn, and R. Michel. Statistik in Theorie und Praxis - mit Anwendungen in R. Springer, Munich, 2014.
M. Falk, F. Marohn, and B. Tewes. Foundations of Statistical Analyses - Examples with SAS. Birkhäuser, Basel, 2003.
W. Gersten. Zielgruppenselektion fĂ¼r Direktmarketingkampagnen. Lang, Frankfurt am Main, 2005.
S. Gross and R. Tibshirani. Data shared lasso: A novel tool to discover uplift. Computational Statistics and Data Analytics, 101:226–235, 2016.
L. Guelman, M. Guillén, and A.M. Perez-Marin. Optimal personalized treatment rules for marketing interventions: A review of methods, a new proposal, and an insurance case study. UB Riskcenter Working Paper Series, 2014(06), 2014.
L. Guelman, M. Guillén, and A.M. Perez-Marin. Uplift random forests. Cybernetics and Systems, (46(3-4)):230–248, 2015.
J. Han and M. Kamber. Data Mining: Concepts and Techniques. Morgan Kaufmann, Elsevier, San Francisco, 2006.
B. Hansotia and B. Rukstales. Incremental value modeling. Journal of Interactive Marketing, 16(3):35–46, 2002.
K. Imai and M. Ratkovic. Estimating treatment effect heterogeneity in randomized program evaluation. The Annals of Applied Statistics, 7(1):443–470, 2013.
M. Jaskowski and S. Jaroszewicz. Uplift modeling for clinical trial data. ICML 2012 Workshop on Clinical Data Analysis, 2012.
F. Kuusisto, V. Costa, H. Nassif, E. Burnside, D. Page, and J. Shavlik. Support vector machines for differential prediction. Proceedings of the 2014th European Conference on Machine Learning and Knowledge Discovery in Databases, 2:50–65, 2014.
K. Larsen. Net models. 2009. Presentation at the 12th Annual Data Mining Conference (2009), available at: http://www.youtube.com/watch?v=JN3WE8IZNVY.
K. Larsen. Net lift models. 2010. Presentation at the Analytics 2010 Conference, available at: http://www.sas.com/events/aconf/2010/pres/larsen.pdf.
K. Larsen. Net Lift Models: Optimizing the Impact of Your Marketing Efforts - Course Notes. SAS Institute Inc., Cary, 2010.
V. Lo. The true lift model - a novel data mining approach to response modeling in database marketing. SIGKDD Explorations, 4(2):78–86, 2002.
R. Michel, I. Schnakenburg, and T. von Martens. Methods of variable pre-selection for netscore modeling. Journal of Research in Interactive Marketing, 7(4):257–268, 2013.
R. Michel, I. Schnakenburg, and T. von Martens. A modified χ 2-test for uplift models with applications in marketing performance measurement. 2014. arXiv:1401.7001.
E.W.T. Ngai, L. Xiu, and D.C.K. Chau. Application of data mining techniques in customer relationship management - a literature review and classification. Expert Systems with Applications, 36(2):2592–2602, 2009.
N.J. Radcliffe. Using control groups to target on predicted lift: Building and assessing uplift models. Direct Marketing Journal, 1:14–21, 2007.
N.J. Radcliffe and P.D. Surry. Real-world uplift modeling with significance-based uplift trees. 2011. Technical Report, Stochastic Solutions.
J. Robbins. A new approach to causal inference in mortality studies with a sustained exposure period-application to control of the healthy worker survivor effect. Mathematical modelling, 7:1395–1512, 1986. https://doi.org/10.1109/ICDM.2010.62.
P.R. Rosenbaum and D.B. Rubin. The central role of the propensity score in observational studies for causal effects. Biometrika, 70:41–55, 1983.
P.R. Rosenbaum and D.B. Rubin. Reducing bias in observational studies using subclassifications on the propensity score. Journal of the American Statistical Association, 79:516–524, 1984.
K. Rudas and S. Jaroszewicz. Linear regression for uplift modeling. Data Mining and Knowledge Discovery, 32:1275–1305, 2018.
P. Rzepakowski and S. Jaroszewicz. Decision trees for uplift modeling with single and multiple treatments. Knowledge and Information Systems, 32:303–327, 2012.
SAS. Statistics 2: ANOVA and Regression Course Notes. SAS Institute Inc., Cary, 2012.
SAS. Advanced Analytics Using SAS Enterprise Miner. SAS Institute Inc., Cary, 2016.
R. Schinazi. Probability with Statistical Applications. Birkhäuser, second edition, 2012.
J.M. Snowden, S. Rose, and K.M. Mortimer. Implementation of G-computation on a simulated data set: Demonstration of a causal inference technique. American Journal of Epidemiology, 173:731–738, 2011.
H. Strasser and C. Weber. On the asymptotic theory of permutation statistics. Mathematical Methods of Statistics, 8:220–250, 1999.
L. Tian, A.A. Alizadeh, A.J. Gentles, and R. Tibshirani. A simple method for detecting interactions between a treatment and a large number of covariates. Journal of the American Statistical Association, 109(508), 2014.
R. Tibshirani. Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society, Series B, 58(1):267–288, 1996.
T. Vanderweele. Explanation in Causal Inference: Methods for Mediation and Interaction. Oxford, 2015.
Y. Zhao, X. Fang, and D. Simchi-Levi. A practically competitive and provably consistent algorithm for uplift modeling. 2017 IEEE International Conference on Data Mining (ICDM), pages 1171–1176, 2017.
Y. Zhao, X. Fang, and D. Simchi-Levi. Uplift modeling with multiple treatments and general response types. Proceedings of the 2017 SIAM International Conference on Data Mining, pages 588–596, 2017.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this chapter
Cite this chapter
Michel, R., Schnakenburg, I., von Martens, T. (2019). Basic Net Scoring Methods: The Uplift Approach. In: Targeting Uplift. Springer, Cham. https://doi.org/10.1007/978-3-030-22625-1_3
Download citation
DOI: https://doi.org/10.1007/978-3-030-22625-1_3
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-22624-4
Online ISBN: 978-3-030-22625-1
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)