Abstract
Predictive performance in model selection is often estimated using out-of-sample validation and test datasets. The assumption is that the test and validation datasets are from the same population as the training dataset. This assumption may not apply in the common application context where the model is applied to scoring of future data. This paper proposes a sample design which can lead to better model performance and robust estimates of model generalization error. The sample design is shown applied to a collection scoring application.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Breiman, L.: Statistical modeling: The two cultures. Statistical Sciences 16(3), 199–231 (2001)
Hand, D., Mannila, H., Smyth, P.: Principles of data mining. MIT Press, Cambridge (2001)
Han, J., Kimber, M.: Data Mining: Concepts and techniques. Morgan Kaufmann, San Francisco (2001)
Witten, I., Frank, E.: Data mining: Machine Learning Tools and Techniques. Morgan Kaufmann, San Francisco (2005)
Maindonald, J.: The role of models in predictive validation (statistics for budding data miners). In: ISIS (2003)
Dhar, V., Stein, R.: Finding robust and usable models from data mining. In: PCAI (1998)
Zadrozny, B.: Learning and Evaluating Classifiers under Sample Selection Bias (2004)
Elkan, C.: Foundations of cost-sensitive learning (2001)
Blake, C., Keogh, E., Merz, C.: Uci repository of machine learning databases. UCI website (2001)
Hand, D.: Classifier technology and the illusion of progress. Statistical Science 21, 1–15 (2006)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2006 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Baxter, R.A. (2006). Finding Robust Models Using a Stratified Design. In: Sattar, A., Kang, Bh. (eds) AI 2006: Advances in Artificial Intelligence. AI 2006. Lecture Notes in Computer Science(), vol 4304. Springer, Berlin, Heidelberg. https://doi.org/10.1007/11941439_123
Download citation
DOI: https://doi.org/10.1007/11941439_123
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-49787-5
Online ISBN: 978-3-540-49788-2
eBook Packages: Computer ScienceComputer Science (R0)