Issues in Stacked Generalization

Abstract

Stacked generalization is a general method of using a high-level model to combine lower-level models to achieve greater predictive accuracy. In this paper we address two crucial issues which have been considered to be a `black art' in classification tasks ever since the introduction of stacked generalization in 1992 by Wolpert: the type of generalizer that is suitable to derive the higher-level model, and the kind of attributes that should be used as its input. We find that best results are obtained when the higher-level model combines the confidence (and not just the predictions) of the lower-level ones. We demonstrate the effectiveness of stacked generalization for combining three different types of learning algorithms for classification tasks. We also compare the performance of stacked generalization with majority vote and published results of arcing and bagging.

Cited by in corpus (14)

• Using stacking to average Bayesian predictive distributions
• The Heterogeneous Ensembles of Standard Classification Algorithms (HESCA): the Whole is Greater than the Sum of its Parts
• Predicting Solar Flares Using CNN and LSTM on Two Solar Cycles of Active Region Data
• A Hybrid Ensemble Learning Approach to Star-Galaxy Classification
• Stacked Penalized Logistic Regression for Selecting Views in Multi-View Learning
• Stacking machine learning classifiers to identify Higgs bosons at the LHC
• Automated Issue Assignment: Results and Insights from an Industrial Case
• Stacking Neural Network Models for Automatic Short Answer Scoring
• Not Hydro: Using Neural Networks to estimate galaxy properties on a Dark-Matter-Only simulation
• View selection in multi-view stacking: Choosing the meta-learner
• ODTE -- An ensemble of multi-class SVM-based oblique decision trees
• Ensemble Quantile Classifier
• Evolving Non-linear Stacking Ensembles for Prediction of Go Player Attributes
• Predicting gestational age at birth in the context of preterm birth from multi-modal fetal MRI