Hostname: page-component-7c8c6479df-ws8qp Total loading time: 0 Render date: 2024-03-17T20:36:50.165Z Has data issue: false hasContentIssue false

What Can We Learn from Predictive Modeling?

Published online by Cambridge University Press:  24 April 2017

Skyler J. Cranmer*
Affiliation:
The Ohio State University, Department of Political Science, 154 North Oval Mall, Columbus, OH 43210, USA. Email: cranmer.12@osu.edu
Bruce A. Desmarais
Affiliation:
Pennsylvania State University, Department of Political Science, 231 Pond Lab, State College, PA 16801, USA. Email: bdesmarais@psu.edu

Abstract

The large majority of inferences drawn in empirical political research follow from model-based associations (e.g., regression). Here, we articulate the benefits of predictive modeling as a complement to this approach. Predictive models aim to specify a probabilistic model that provides a good fit to testing data that were not used to estimate the model’s parameters. Our goals are threefold. First, we review the central benefits of this under-utilized approach from a perspective uncommon in the existing literature: we focus on how predictive modeling can be used to complement and augment standard associational analyses. Second, we advance the state of the literature by laying out a simple set of benchmark predictive criteria. Third, we illustrate our approach through a detailed application to the prediction of interstate conflict.

Type
Articles
Copyright
Copyright © The Author(s) 2017. Published by Cambridge University Press on behalf of the Society for Political Methodology. 

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Footnotes

Authors’ note: Many thanks to Alison Craig for research assistance. Sincere thanks also to Matt Blackwell and Michael Neblo for helpful comments on an earlier draft. The authors are grateful for the support of the National Science Foundation (SES-1558661, SES-1619644, SES-1637089, CISE-1320219, SES-1357622, SES-1514750, and SES-1461493) and the Alexander von Humboldt Foundation. Replication data are posted to the Political Analysis Dataverse (Cranmer and Desmarais 2016a).

Contributing Editor: Jonathan Katz

References

Achen, Christopher H. 2002. Toward a new political methodology: Microfoundations and ART. Annual Review of Political Science 5(1):423450.Google Scholar
Adamic, Lada A., and Adar, Eytan. 2003. Friends and neighbors on the web. Social Networks 25(3):211230.Google Scholar
Airoldi, Edoardo M., Blei, David M., Fienberg, Stephen E., and Xing, Eric P.. 2008. Mixed membership stochastic blockmodels. Journal of Machine Learning Research 9:19812014.Google Scholar
Attewell, Paul, Monaghan, David B., and Kwong, Darren. 2015. Preparing training and test datasets . 1 edn. University of California Press, pp. 6371. http://www.jstor.org/stable/10.1525/j.ctt13x1gcg.8.Google Scholar
Beck, Nathaniel, Katz, Jonathan N., and Tucker, Richard. 1998. Taking time seriously: Time-series-cross-section analysis with a binary dependent variable. American Journal of Political Science 42(4):12601288.Google Scholar
Beck, Nathaniel, King, Gary, and Zeng, Langche. 2000. Improving quantitative studies of international conflict: A conjecture. American Political Science Review 94(1):2135.Google Scholar
Brandt, Patrick T., Freeman, John R., and Schrodt, Philip A.. 2011. Real time, time series forecasting of political conflict. Conflict Management and Peace Science 28(1):4164.Google Scholar
Cawley, Gavin C., and Talbot, Nicola L. C.. 2010. On over-fitting in model selection and subsequent selection bias in performance evaluation. The Journal of Machine Learning Research 11:20792107.Google Scholar
Clarke, Kevin A., and Primo, David M.. 2007. Modernizing political science: A model-based approach. Perspectives on Politics 5(4):741753.Google Scholar
Collopy, Fred, Adya, Monica, and Armstrong, J. Scott. 1994. Principles for examining predictive validity – The case of information systems spending forecasts. Information Systems Research 5(2):170179.Google Scholar
Cranmer, Skyler J., and Desmarais, Bruce A.. 2011. Inferential network analysis with exponential random graph models. Political Analysis 19(1):6686.Google Scholar
Cranmer, Skyler, and Desmarais, Bruce. 2016a. Replication data for: What can we learn from predictive modeling? http://dx.doi.org/10.7910/DVN/UFSQ1J, Harvard Dataverse.Google Scholar
Cranmer, Skyler J., and Desmarais, Bruce A.. 2016b. A critique of dyadic design. International Studies Quarterly 60(2):355362.Google Scholar
Cranmer, Skyler J., Desmarais, Bruce A., and Kirkland, Justin H.. 2012. Towards a network theory of alliance formation. International Interactions 38(3):295324.Google Scholar
Cranmer, Skyler J., Desmarais, Bruce A., and Menninga, Elizabeth J.. 2012. Complex dependencies in the alliance network. Conflict Management and Peace Science 29(3):279313.Google Scholar
Cranmer, Skyler J., Menninga, Elizabeth J., and Mucha, Peter J.. 2015. Kantian fractionalization predicts the conflict propensity of the international system. Proceedings of the National Academy of Sciences 112(38):1181211816.Google Scholar
Cranmer, Skyler J., Rice, Douglas, and Siverson, Randolph M.. 2015. What to do about atheoretic lags. Political Science Research Methods, doi:10.1017/psrm.2015.36.Google Scholar
Cybenko, George. 1989. Approximation by superpositions of a sigmoidal function. Mathematics of Control, Signals and Systems 2(4):303314.Google Scholar
Dafoe, Allan. 2011. Statistical critiques of the democratic peace: Caveat emptor. American Journal of Political Science 55(2):247262.Google Scholar
Davis, Jesse, and Goadrich, Mark. 2006. The relationship between precision-recall and ROC curves. In Proceedings of the 23rd International Conference on Machine Learning, ICML ’06 . New York, NY: ACM, pp. 233240. http://doi.acm.org/10.1145/1143844.1143874.Google Scholar
Desmarais, Bruce A., and Cranmer, Skyler J.. 2010. Consistent confidence intervals for maximum pseudolikelihood estimators. In Proceedings of the Neural Information Processing Systems 2010 Workshop on Computational Social Science and the Wisdom of Crowds .Google Scholar
Desmarais, Bruce A., and Cranmer, Skyler J.. 2011. Forecasting the locational dynamics of transnational terrorism: A network analytic approach. In Proceedings of the European Intelligence and Security Informatics Conference (EISIC) 2011 , Athens, Greece: IEEE Computer Society.Google Scholar
Desmarais, Bruce A., and Cranmer, Skyler J.. 2012. Statistical mechanics of networks: Estimation and uncertainty. Physica A 391(4):18651876.Google Scholar
Dettling, Marcel, and Bühlmann, Peter. 2003. Boosting for tumor classification with gene expression data. Bioinformatics 19(9):10611069.Google Scholar
Droge, Bernd. 1999. Asymptotic optimality of full cross-validation for selecting linear regression models. Statistics and Probability Letters 44(4):351357.Google Scholar
Druckman, James N., Green, Donald P., Kuklinski, James H., and Lupia, Arthur. 2006. The growth and development of experimental research in political science. American Political Science Review 100(04):627635.Google Scholar
Esteban, Cristóbal, Schmidt, Danilo, Krompaß, Denis, and Tresp, Volker. 2015. Predicting sequences of clinical events by using a personalized temporal latent embedding model. In Healthcare Informatics (ICHI), 2015 International Conference on IEEE , pp. 130139.Google Scholar
Faraway, Julian James. 2006. Extending the linear model with R: Generalized linear, mixed effects and nonparametric regression models, vol. 66 . New York: CRC press.Google Scholar
Fawcett, Tom. 2006. An introduction to ROC analysis. Pattern Recognition Letters 27(8):861874.Google Scholar
Gartzke, Erik. 2007. The capitalist peace. American Journal of Political Science 51(1):166191.Google Scholar
Gill, Jeff. 2014. Bayesian methods: A social and behavioral sciences approach . Boca Raton: Chapman and Hall/CRC.Google Scholar
Gleditsch, Kristian Skrede. 2002. Expanded trade and GDP data. Journal of Conflict Resolution 46(5):712724.Google Scholar
Gleditsch, Kristian S., and Ward, Michael D.. 2001. Measuring space: A minimum-distance database and applications to international studies. Journal of Peace Research 38(6):739758.Google Scholar
Goeman, J. J., Meijer, R. J., and Chaturvedi, N.. 2016. Penalized: L1 (lasso and fused lasso) and L2 (ridge) penalized estimation in GLMs and in the Cox model. R package version 0.9-46. https://cran.r-project.org/web/packages/penalized/index.html.Google Scholar
Goldstone, Jack A., Bates, Robert H., Epstein, David L., Gurr, Ted Robert, Lustik, Michael B., Marshall, Monty G., Ulfelder, Jay, and Woodward, Mark. 2010. A global model for forecasting political instability. American Journal of Political Science 54(1):190208.Google Scholar
Gurbaxani, Vijay, and Mendelson, Haim. 1990. An integrative model of information systems spending growth. Information Systems Research 1(1):2346.Google Scholar
Gurbaxani, Vijay, and Mendelson, Haim. 1994. Modeling vs. forecasting—The case of information systems spending. Information Systems Research 5(2):180190.Google Scholar
Hall, Peter. 1983. Large sample optimality of least squares cross-validation in density estimation. The Annals of Statistics 11(4):11561174.Google Scholar
Hanneke, Steve, Fu, Wenjie, and Xing, Eric P.. 2010. Discrete temporal models of social networks. The Electronic Journal of Statistics 4:585605.Google Scholar
Hastie, Trevor, Tibshirani, Robert, and Friedman, Jerome. 2009. The elements of statistical learning: Data mining, inference, and prediction . 2nd edn. New York: Springer.Google Scholar
Hoff, Peter D., Raftery, Adrian E., and Handcock, Mark S.. 2002. Latent space approaches to social network analysis. Journal of the American Statistical association 97(460):10901098.Google Scholar
Hua, Wang, Cuiqin, Ma, and Lijuan, Zhou. 2009. A brief review of machine learning and its application. In Information engineering and computer science, 2009. ICIECS 2009. International Conference on IEEE , pp. 14.Google Scholar
Jensen, David D., and Cohen, Paul R.. 2000. Multiple comparisons in induction algorithms. Machine Learning 38(3):309338.Google Scholar
Keele, Luke. 2015. The statistics of causal inference: A view from political methodology. Political Analysis 23:313335.Google Scholar
Kuhn, Max, and Johnson, Kjell. 2013. Applied predictive modeling . New York: Springer.Google Scholar
van der Laan, Mark J., Dudoit, Sandrine, and Keles, Sunduz. 2004. Asymptotic optimality of likelihood-based cross-validation. Statistical Applications in Genetics and Molecular Biology 3(1):123.Google Scholar
Leicht, Elizabeth A., Holme, Petter, and Newman, Mark E. J.. 2006. Vertex similarity in networks. Physical Review E 73(2):026120.Google Scholar
Lopes, Miguel, and Bontempi, Gianluca. 2014. On the null distribution of the precision and recall curve. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases Berlin: Springer, pp. 322337.Google Scholar
Marshall, Monty G., and Jaggers, Keith. 2002. Polity IV project: Political regime characteristics and transitions, pp. 1800–2002. http://www.systemicpeace.org/polityproject.html.Google Scholar
Meyer, Patrick E., Lafitte, Frederic, and Bontempi, Gianluca. 2008. MINET: An open source R/Bioconductor package for mutual information based network inference. BMC Bioinformatics 9.Google Scholar
Muchlinski, David, Siroky, David, He, Jingrui, and Kocher, Matthew. 2016. Comparing random forest with logistic regression for predicting class-imbalanced civil war onset data. Political Analysis 24(1):87103.Google Scholar
Nadeau, Claude, and Bengio, Yoshua. 2003. Inference for the generalization error. Machine Learning 52(3):239281.Google Scholar
Nowak, Robert D. 1997. Optimal signal estimation using cross-validation. IEEE Signal Processing Letters 4(1):2325.Google Scholar
Oneal, John, and Russett, Bruce M.. 1999. The Kantian peace: The Pacific benefits of democracy, interdependence, and international organization. World Politics 52(1):137.Google Scholar
Oneal, John R., and Russett, Bruce. 2005. Rule of three, let it be? When more really is better. Conflict Management and Peace Science 22(4):293310.Google Scholar
Ozenne, Brice, Subtil, Fabien, and Maucort-Boulch, Delphine. 2015. The precision–recall curve overcame the optimism of the receiver operating characteristic curve in rare diseases. Journal of Clinical Epidemiology 68(8):855859.Google Scholar
Pevenhouse, Jon C., and Goldstein, Joshua S.. 1999. Serbian compliance or defiance in Kosovo? Statistical analysis and real-time predictions. Journal of Conflict Resolution 43(4):538546.Google Scholar
Pevehouse, Jon, Nordstrom, Timothy, and Warnke, Kevin. 2004. The correlates of war 2 international governmental organizations data version 2.0. Conflict Management and Peace Science 21(2):101119.Google Scholar
Pons, Pascal, and Latapy, Matthieu. 2005. Computing communities in large networks using random walks. Journal of Graph Algorithms and Applications 10(2):191218.Google Scholar
Rakotomalala, Ricco, Chauchat, Jean-Hughes, and Pellegrino, Francois. 2006. Accuracy estimation with clustered dataset. In Proceedings of the Fifth Australasian Conference on Data Mining and Analystics 61:17–22 . Sydney, Australia: Australian Computer Society, Inc.Google Scholar
Ripley, Brian, and Venables, William. 2016. Nnet: Feed-forward neural networks and multinomial log-linear models ., R package version 7.3-12. https://cran.r-project.org/web/packages/nnet.Google Scholar
Rost, Nicholas, Schneider, Gerald, and Kleibl, Johannes. 2009. A global risk assessment model for civil wars. Social Science Research 38(4):921933.Google Scholar
Schneider, Gerald. 2012. Banking on the Broker: Forecasting conflict in the levant with financial data. In Illuminating the shadow of the future: scientific prediction and the human condition , ed. Wayman, Frank, Williamson, Paul, and Bueno de Mesquita, Bruce. Ann Arbor: University of Michigan Press.Google Scholar
Schneider, Gerald, Gleditsch, Nils Petter, and Carey, Sabine. 2010. Exploring the past, anticipating the future: A symposium. International Studies Review 12(1):17.Google Scholar
Schneider, Gerald, Gleditsch, Nils Petter, and Carey, Sabine. 2011. Forecasting in international relations: One quest, three approaches. Conflict Management and Peace Science 28(5):514.Google Scholar
Schrodt, Philip A., and Gerner, Deborah J.. 2000. Using cluster analysis to derive early warning indicators for political change in the middle east, 1979–1996. American Political Science Review 94(4):803818.Google Scholar
Shmueli, Galit. 2010. To explain or to predict?. Statistical Science 25(3):289310.Google Scholar
Sing, T., Sander, O., Beerenwinkel, N., and Lengauer, T.. 2005. ROCR: Visualizing classifier performance in R. Bioinformatics 21(20):7881.Google Scholar
Singer, J. David, Bremer, Stuart, and Stuckey, John. 1972. Capability distribution, uncertainty, and major power war, 1820–1965. Peace, war, and numbers 19:48.Google Scholar
Stinnett, Douglas M., Tir, Jaroslav, Diehl, Paul F., Schafer, Philip, and Gochman, Charles. 2002. The Correlates of War (COW) project direct contiguity data, version 3.0. Conflict Management and Peace Science 19(2):5967.Google Scholar
Stone, M. 1974. Cross-validatory choice and assessment of statistical predictions. Journal of the Royal Statistical Society, Series B 36(2):111147.Google Scholar
Stone, M. 1977. Asymptotics for and against cross-validation. Biometrika 64(1):2935.Google Scholar
Tuszynski, Jarek. 2014. caTools: Tools: Moving window statistics, GIF, Base64, ROC AUC, etc. R package version 1.17. http://CRAN.R-project.org/package=caTools.Google Scholar
Van Maanen, John, Sørensen, Jesper B., and Mitchell, Terence R.. 2007. The interplay between theory and method. Academy of Management Review 32(4):11451154.Google Scholar
Ward, Michael D., Greenhill, Brian D., and Bakke, Kristin M.. 2010. The perils of policy by p-value: Predicting civil conflicts. Journal of Peace Research 47(4):363375.Google Scholar
Ward, Michael D., Siverson, Randolph M., and Cao, Xun. 2007. Disputes, democracies, and dependencies: A reexamination of the Kantian peace. American Journal of Political Science 51(3):583601.Google Scholar
Zou, Hui, and Hastie, Trevor. 2005. Regularization and variable selection via the elastic net. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 67(2):301320.Google Scholar