Using Model Trees for Classification

By Eibe Frank, Yong Wang, Stuart Inglis, G. Holmes, Ian H. Witten. 1998

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. David López Vilariño
2. R. Fisher
3. Eibe Frank
4. Sina Aghaei
5. João Gama
6. Shilong Piao
7. D. Lawrence
8. C. Koven

Citations in the corpus

• Learning Optimal and Fair Decision Trees for Non-Discriminative Decision-Making (2019, methodology): “An approximate algorithm for designing classification trees with linear leafing rules was originally proposed in (Frank et al. 1998).”
• Exploiting place features in link prediction on location-based social networks (2011, background): “5[20]), Näıve Bayes, model trees with linear regression on the leaves [10], and random forests (10 trees, 4 random features each) [24].”
• A framework for benchmarking land models (2012, background): “That is, the available data streams are insufficient to constrain model parameterization (Weng and Luo, 2011; Wang et al., 2001; Carvalhais et al., 2010) or to distinguish between different modeling structures (Frank et al., 1998).”
• Is Combining Classifiers with Stacking Better than Selecting the Best One? (2004, background): “The results of Frank et al. (1998), who investigate classification via regression, show that classification via model trees performs extremely well, i.”
• Wind power forecasting : state-of-the-art 2009. (2009, methodology): “in [132] tested five different data-mining models [133] [134] to forecast the wind power: SVM, MLP NN, the M5P tree algorithm, the Reduced Error Pruning tree, and the bagging tree.”
• Functional Trees (2001, methodology): “For comparative purposes we also evaluated the state of the art in univariate trees, C4.5 (Quinlan, 1993a), M5 Class (Witten & Frank, 2000), and the univariate and multivariate versions of Cruise (Kim & Loh, 2001).”
• Graph-based approach for airborne light detection and ranging segmentation (2014, methodology): “ClassificationViaRegression w (Frank et al., 1998) binarizes each class and learns its corresponding M5P tree/rule regression model (Quinlan, 1992), with at least 4 training patterns per leaf.”
• Logistic Model Trees (2003, methodology): “We also include results for two learning schemes that build multiple trees, namely boosted decision trees and model trees fit to the class indicator variables [4], and a different algorithm for building logistic model trees called PLUS [8].”
• A Survey of Evolutionary Algorithms for Decision-Tree Induction (2012, unknown): “[106], [122], REPTree [123] and OC1 [117].”