作者: Francisco Moreno-Seco , José M. Iñesta , Pedro J. Ponce de León , Luisa Micó
DOI: 10.1007/11815921_77
关键词: Ensembles of classifiers 、 Artificial intelligence 、 Computer science 、 Random subspace method 、 Support vector machine 、 Cascading classifiers 、 Pattern recognition 、 Classifier (UML) 、 Sensor fusion 、 Voting
摘要: This work presents a comparison of current research in the use voting ensembles classifiers order to improve accuracy single and make performance more robust against difficulties that each individual classifier may have. Also, number combination rules are proposed. Different schemes discussed compared study ensemble task. The have been trained on real data available for benchmarking also applied case related statistical description models melodies music genre recognition.