Using ensembles for problems with characterizable changes in data distribution: A case study on quantification
详细信息 查看全文
- 作者：Pablo Pé ; rez-Gá ; llego pablopg@aic.uniovi.es" class="auth_mail" title="E-mail the corresponding author ; José ; Ramó ; n Quevedo quevedo@aic.uniovi.es" class="auth_mail" title="E-mail the corresponding author ; Juan José ; del Coz ; juanjo@aic.uniovi.es" class="auth_mail" title="E-mail the corresponding author ; juanjo@uniovi.es" class="auth_mail" title="E-mail the corresponding author
- 关键词：Distribution changes ; Ensembles ; Quantification
- 刊名：Information Fusion
- 出版年：2017
- 出版时间：March 2017
- 年：2017
- 卷：34
- 期：Complete
- 页码：87-100
- 全文大小：1036 K
文摘
- class="listitem" id="list_celist0001">
- class="label">•
Our hypothesis: ensembles are well suited for problems with distribution changes.
- class="label">•
If those changes are characterizable, ensembles can be designed to tackle them.
- class="label">•
Idea: to generate different samples based on the expected distribution changes.
- class="label">•
Case study: we present ensembles versions of two binary quantification algorithms.
- class="label">•
Ensembles outperform original counterpart algorithms using trivial aggregation rules.