| 翻訳と辞書 |
| Cluster Based Ensemble Classifiers ： ウィキペディア英語版 |
Cluster Based Ensemble Classifiers
Cluster Based Ensemble Classifiers
Ensemble classifiers, part of the machine learning branch of artificial intelligence, refers to a group of classifiers that collectively learn the decision boundaries between classes. The idea is to train multiple classifiers in a complementary fashion so that weakness/mistakes of one base classifier is compensated by another. Data can be partitioned into multiple clusters and classifiers can be trained on them. This is the concept of cluster based ensemble classifier. With hard clustering, however, patterns do not overlap between clusters and one base classifier cannot compensate for errors made by another base classifier.
There is a new set of research works aims to achieve this overlapping 〔A. Rahman and B. Verma, Cluster Based Ensemble of Classifiers, Wiley Expert Systems, DOI: 10.1111/j.1468-0394.2012.00637.x, 2012.〕〔B. Verma and A. Rahman, Cluster Oriented Ensemble Classifier: Impact of Multi-cluster Characterisation on Ensemble Classifier Learning, IEEE Transaction on Knowledge and Data Engineering, vol. 24, no. 4, pp. 605–618, 2012.〕〔A. Rahman and B. Verma, A Novel Layered Clustering based Approach for Generating Ensemble of Classifiers, IEEE Transaction on Neural Networks, vol 22, no 5, pp 781–792, 2011.〕〔A. Rahman and B. Verma, Cluster Oriented Ensemble Classifiers using Multi–Objective Evolutionary Algorithm, Proc. IEEE International Joint Conference on Neural Networks (IJCNN), pp. 829–834, Dallas, Texas, 2013.〕〔A. Rahman and B. Verma, Influence of Unstable Patterns in Layered Cluster Oriented Ensemble Classifier,” Proc. IEEE International Joint Conference on Neural Networks (IJCNN), Brisbane, Australia, 2012.〕〔A. Rahman, B. Verma, and X. Yao, Non–uniform Layered Clustering for Ensemble Classifier Generation and Optimality, 19th International Conference on Neural Information Processing (ICONIP 2012): Lecture Notes in Computer Science, vol. 6443, pp. 551–558, 2010.〕〔A. Rahman and B. Verma, A Novel Ensemble Classifier Approach using Weak Classifier Learning on Overlapping Clusters, Proc. IEEE International Joint Conference on Neural Networks (IJCNN), Barcelona, Spain, 2010.〕 The works aim at clustering data at multiple layers, and train base classifiers on the clusters at each layer. Patterns belong to different clusters at different layers and classifiers trained at one layers can compensate for the errors in another layer. Classification accuracy achieved by layered cluster oriented ensemble classifiers is better than bagging and boosting as presented in the papers.
==References==
抄文引用元・出典: フリー百科事典『 ウィキペディア（Wikipedia）』
■ウィキペディアで「Cluster Based Ensemble Classifiers」の詳細全文を読む
スポンサード リンク
| 翻訳と辞書 : 翻訳のためのインターネットリソース |
Copyright(C) kotoba.ne.jp 1997-2016. All Rights Reserved.
|
|