Abstract

Interval prediction can be more useful than single value prediction in many continuous data streams. This paper introduces a novel Interval Prediction Tree IP3 algorithm for interval prediction of numerical target variables from temporal mean-variance aggregated continuous data. This algorithm characterized by: processing incoming mean-variance aggregated multivariate temporal data, splitting each of the continuous features of the input according to the best mean-variance and making stable interval predictions of a target numerical variable with a given degree of statistical confidence. As shown by empirical evaluations in forest fires data set the proposed method provides better performance than existing regression tree models.

References

• Alberg, D., Last, M., & Kandel, A. (2012). Knowledge Discovery in Data Streams with Regression Tree Methods. WIREs Data Mining Knowledge Discovery 2012 , 2, 69-78.
• Bishop, C. (2006). Pattern Recognition and Machine Learning. New York: Springer.
• Breiman, L., & Friedman, J. (1985). Estimating Optimal Transformations for Multiple Regression and Correlation. Journal of American Statistic Association , 80, 580 - 597.
• Breiman, L., Friedman, J., Olshen, R., & Stone, C. (1984). Classification and Regression Trees. Belmont: Wadsworth & Brooks/Cole. Pacific Grove.
• Ceci, M., Appice, A., & Malerba, D. (2003). Comparing Simplification Methods for Model Trees with Regression and Splitting Nodes. Foundations of Intelligent Systems, 14th International LNAI Symposium, 2871, pp. 49 - 56.
• Cortez, P., & Morais, A. (2007). A Data Mining Approach to Predict Forest Fires using Meteorological Data. In M. F. In J. Neves (Ed.), New Trends in Artificial Intelligence, Proceedings of the 13th EPIA 2007 - Portuguese Conference on Artificial Intelligence, (pp. 512-523). Guimaraes, Portugal.
• Friedman, J. (1991). Multivariate Adaptative Regression Splines. In Annals of Statistics , 1 - 19.
• Friedman, J. (2002). Stochastic Gradient Boosting. Computational Statistics & Data Analysis , 38 (4), 367 - 378.
• Ikonomovska, E., Gama, J., Sebastiao, R., & Gjorgjevik, D. (2009). Regression Trees from Data Streams with Drift Detection. Discovery Science, (pp. 121 - 135).
• Karalic, A. (1992). Linear Regression in Regression Tree Leaves. In Proceedings of International School for Synthesis of Expert Knowledge, 10(3), pp. 151 - 162.
• Loh, W. (2009). Regression by Parts: Fitting Visually Interpretable Models with GUIDE. (W. H. in C. Chen, Ed.)
• Quinlan, J. (1992). Learning with Continuous Classes. In Proceedings of the 5th Australian Joint Conference on Artificial Intelligence. World Scientific.
• Vens, C., & Blockeel, H. (2006). A Simple Regression Based Heuristic for Learning Model Trees. Intelligent Data Analysis , 10(3), 215 - 236.
• Wang, Y., & Witten, I. (1997). Inducing of Model Trees for Predicting Continuous Classes. In Proceedings of the 9th European Conference on Machine Learning (pp. 128 - 137). Springer-Verlag.