@inproceedings{f5c8ce3580a54c26bfc38caadfadd353,
title = "A fuzzy integral method of applying support vector machine for multi-class problem",
abstract = "This paper proposed a novel method of applying support vector machine for multi-class problem based on fuzzy integral. Firstly, the fuzzy measure of each binary classifier is constructed based on its classification accuracy during training and its agreement degrees to other support vector machines. Then the testing instances are classified by calculating the fuzzy integral between the fuzzy measures and the outputs of the binary support vector machines. The experiment results on iris and glass datasets from UCI machine learning repository and real plane dataset show that the new method is effective. And the experiment results ulteriorly indicate that the method with Choquet fuzzy integral has better performance than that with Sugeno integral.",
author = "Yanning Zhang and Hejin Yuan and Jin Pan and Ying Li and Runping Xi and Lan Yao",
year = "2006",
doi = "10.1007/11881223_107",
language = "英语",
isbn = "3540459073",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "839--846",
booktitle = "Advances in Natural Computation - Second International Conference, ICNC 2006, Proceedings",
note = "2nd International Conference on Natural Computation, ICNC 2006 ; Conference date: 24-09-2006 Through 28-09-2006",
}