372 CHAPTER8 Ensemble Learning Rodriguez (2007) show that the main factors responsible for its performance are the use of principal components transformations (as opposed to other feature-extraction methods such as random projections) and the application of principal components analysis to random subspaces of the original input attributes. Freund and Schapire (1996) developed the AdaBoost.M1 boosting algorithm and derived theoretical bounds for its performance. Later they improved these bounds using the concept of margins (Freund and Schapire, 1999). Drucker (1997) adapted AdaBoost.M1 for numeric prediction. The LogitBoost algorithm was devel- oped by Friedman et al. (2000). Friedman (2001) describes how to make boosting more resilient in the presence of noisy data. Domingos (1997) describes how to derive a single interpretable model from an ensemble using artificial training examples. Bayesian option trees were introduced by Buntine (1992), and majority voting was incorporated into option trees by Kohavi and Kunz (1997). Freund and Mason (1999) introduced alternating decision trees; experiments with multiclass alternating decision trees were reported by Holmes et al. (2002). Landwehr et al. (2005) developed logistic model trees using the LogitBoost algorithm. Stacked generalization originated with Wolpert (1992), who presented the idea