Classifier combination; Committee-based learning; Multiple classifier system
Ensemble is a learning paradigm where multiple learners are trained to solve the same problem. In contrast to ordinary learning approaches that try to learn one hypothesis from training data, ensemble methods try to construct and combine a set of hypotheses.
It is difficult to trace the starting point of the history of ensemble methods since the basic idea of deploying multiple models has been in use for a long time. However, it is clear that the hot wave of research on ensemble methods since the 1990s owes much to two works. The first is an applied research conducted by Hansen and Salamon at the end of 1980s , where they found that predictions made by the combination of a set of neural networks are often more accurate than predictions made by the best single neural network. The second is a theoretical research conducted in 1990, where Schapire proved that weak...
- 2.Schapire RE. The Boosting approach to machine learning: an overview. In: Denison DD, Hansen MH, Holmes C, Mallick B, Yu B, editors. Nonlinear Estimation and Classification. Berlin: Springer; 2003.Google Scholar
- 9.Dietterich TG. Machine learning research: four current directions. AI Magn. 1997;18(4):97–136.Google Scholar
- 12.Krogh A. Neural network ensembles, cross validation, and active learning. In: Tesauro G, Touretzky DS, Leen TK, editors. Advances in Neural Information Processing Systems 7. Cambridge, MA: MIT Press; 1995. p. 231–8.Google Scholar