Publications-Proceedings
Article View/Open
Publication Export
Google ScholarTM
NCCU Library
Citation Infomation
-
No data in Web of Science(Wrong one)Loading...
Related Publications in TAIR
| Title | Integrating adaptive boosting and support vector machines with varying kernels |
| Creator | Hsu, Kuo-Wei 徐國偉 |
| Contributor | 資訊科學系 |
| Key Words | Adaptive boosting; Information management; Iterative methods; Learning algorithms; Support vector machines; Vector spaces; Base learners; Classification algorithm; Classification models; Classification performance; Collective classifications; Multi-class; Multiple kernels; Single kernel; Classification (of information) |
| Date | 2017-01 |
| Date Issued | 3-Aug-2017 14:13:58 (UTC+8) |
| Summary | Adaptive Boosting, or AdaBoost, is a meta-learning algorithm that employs a classification algorithm as a base learner to train classification models and uses these models to perform collective classification. One of its main features is that iteratively it forces the base learner to work more on difficult samples. Usually it can achieve better overall classification performance, when compared to a single classification model trained by the classification algorithm used as the base learner. SVM, short for Support Vector Machine, is a learning algorithm that employs a kernel to project the original data space to a data space where a hyperplane that can linearly separate as many samples of classes as possible can be found. Because both are top algorithms, researchers have been exploring the use of AdaBoost with SVM. Unlike others simply using SVM with a single kernel as the base learner in AdaBoost, we propose an approach that uses SVM with multiple kernels as the base learners in a variant of AdaBoost. Its main feature is that it not only considers difficulties of samples but also classification performance of kernels, and accordingly it selects as well as switches between kernels in the boosting process. The experiment results show that we can obtain better classification performance by using the proposed approach. © 2017 ACM. |
| Relation | Proceedings of the 11th International Conference on Ubiquitous Information Management and Communication, IMCOM 2017, 11th International Conference on Ubiquitous Information Management and Communication, IMCOM 2017; Beppu; Japan; 5 January 2017 到 7 January 2017; 代碼 126221 |
| Type | conference |
| DOI | http://dx.doi.org/10.1145/3022227.3022314 |
| dc.contributor | 資訊科學系 | zh_Tw |
| dc.creator (作者) | Hsu, Kuo-Wei | en_US |
| dc.creator (作者) | 徐國偉 | zh_TW |
| dc.date (日期) | 2017-01 | en_US |
| dc.date.accessioned | 3-Aug-2017 14:13:58 (UTC+8) | - |
| dc.date.available | 3-Aug-2017 14:13:58 (UTC+8) | - |
| dc.date.issued (上傳時間) | 3-Aug-2017 14:13:58 (UTC+8) | - |
| dc.identifier.uri (URI) | http://nccur.lib.nccu.edu.tw/handle/140.119/111628 | - |
| dc.description.abstract (摘要) | Adaptive Boosting, or AdaBoost, is a meta-learning algorithm that employs a classification algorithm as a base learner to train classification models and uses these models to perform collective classification. One of its main features is that iteratively it forces the base learner to work more on difficult samples. Usually it can achieve better overall classification performance, when compared to a single classification model trained by the classification algorithm used as the base learner. SVM, short for Support Vector Machine, is a learning algorithm that employs a kernel to project the original data space to a data space where a hyperplane that can linearly separate as many samples of classes as possible can be found. Because both are top algorithms, researchers have been exploring the use of AdaBoost with SVM. Unlike others simply using SVM with a single kernel as the base learner in AdaBoost, we propose an approach that uses SVM with multiple kernels as the base learners in a variant of AdaBoost. Its main feature is that it not only considers difficulties of samples but also classification performance of kernels, and accordingly it selects as well as switches between kernels in the boosting process. The experiment results show that we can obtain better classification performance by using the proposed approach. © 2017 ACM. | en_US |
| dc.format.extent | 297842 bytes | - |
| dc.format.mimetype | application/pdf | - |
| dc.relation (關聯) | Proceedings of the 11th International Conference on Ubiquitous Information Management and Communication, IMCOM 2017, | en_US |
| dc.relation (關聯) | 11th International Conference on Ubiquitous Information Management and Communication, IMCOM 2017; Beppu; Japan; 5 January 2017 到 7 January 2017; 代碼 126221 | en_US |
| dc.subject (關鍵詞) | Adaptive boosting; Information management; Iterative methods; Learning algorithms; Support vector machines; Vector spaces; Base learners; Classification algorithm; Classification models; Classification performance; Collective classifications; Multi-class; Multiple kernels; Single kernel; Classification (of information) | en_US |
| dc.title (題名) | Integrating adaptive boosting and support vector machines with varying kernels | en_US |
| dc.type (資料類型) | conference | |
| dc.identifier.doi (DOI) | 10.1145/3022227.3022314 | |
| dc.doi.uri (DOI) | http://dx.doi.org/10.1145/3022227.3022314 |