Collaborative SVM classification in scale-free peer-to-peer networks
详细信息 查看全文
- 作者：Umer Khan ; a ; khan@ismll.uni-hildesheim.de" class="auth_mail" title="E-mail the corresponding author ; khan@ismll.de" class="auth_mail" title="E-mail the corresponding author ; Lars Schmidt-Thiemea ; schmidt-thieme@ismll.uni-hildesheim.de" class="auth_mail" title="E-mail the corresponding author ; Alexandros Nanopoulos1 ; b ; alexandros.nanopoulos@ku.de" class="auth_mail" title="E-mail the corresponding author
- 关键词：Distributed ; Classification ; SVM ; P2P ; Skew
- 刊名：Expert Systems with Applications
- 出版年：2017
- 出版时间：1 March 2017
- 年：2017
- 卷：69
- 期：Complete
- 页码：74-86
- 全文大小：2092 K
文摘
- •
Collaborative SVM method in decentralized, hierarchical & skewed P2P networks.
- •
Performs transitive propagation of models beyond the direct neighbors.
- •
Considers vast majority of scarcely connected peers.
- •
Analysis w.r.t music genre classification, centralized methods and data replication.
- •
Improves the overall accuracy substantially and keeps the communication cost low.