1 resultado para active data-centric
em DigitalCommons@University of Nebraska - Lincoln
Filtro por publicador
- KUPS-Datenbank - Universität zu Köln - Kölner UniversitätsPublikationsServer (1)
- Aberdeen University (2)
- Academic Archive On-line (Stockholm University; Sweden) (1)
- AMS Tesi di Dottorato - Alm@DL - Università di Bologna (17)
- AMS Tesi di Laurea - Alm@DL - Università di Bologna (1)
- ArchiMeD - Elektronische Publikationen der Universität Mainz - Alemanha (3)
- Archimer: Archive de l'Institut francais de recherche pour l'exploitation de la mer (4)
- Aston University Research Archive (6)
- Biblioteca Digital da Produção Intelectual da Universidade de São Paulo (11)
- Biblioteca Digital da Produção Intelectual da Universidade de São Paulo (BDPI/USP) (150)
- Biblioteca Virtual del Sistema Sanitario Público de Andalucía (BV-SSPA), Junta de Andalucía. Consejería de Salud y Bienestar Social, Spain (2)
- BORIS: Bern Open Repository and Information System - Berna - Suiça (31)
- Brock University, Canada (3)
- Bulgarian Digital Mathematics Library at IMI-BAS (1)
- CentAUR: Central Archive University of Reading - UK (34)
- CiencIPCA - Instituto Politécnico do Cávado e do Ave, Portugal (1)
- Cochin University of Science & Technology (CUSAT), India (2)
- Coffee Science - Universidade Federal de Lavras (3)
- Consorci de Serveis Universitaris de Catalunya (CSUC), Spain (12)
- CORA - Cork Open Research Archive - University College Cork - Ireland (2)
- CUNY Academic Works (1)
- Dalarna University College Electronic Archive (1)
- Digital Commons - Michigan Tech (6)
- Digital Commons @ DU | University of Denver Research (1)
- Digital Commons at Florida International University (7)
- Digital Peer Publishing (1)
- DigitalCommons@The Texas Medical Center (10)
- DigitalCommons@University of Nebraska - Lincoln (1)
- Doria (National Library of Finland DSpace Services) - National Library of Finland, Finland (15)
- DRUM (Digital Repository at the University of Maryland) (3)
- FUNDAJ - Fundação Joaquim Nabuco (2)
- Instituto Nacional de Saúde de Portugal (1)
- Instituto Politécnico do Porto, Portugal (5)
- Iowa Publications Online (IPO) - State Library, State of Iowa (Iowa), United States (2)
- Martin Luther Universitat Halle Wittenberg, Germany (1)
- Massachusetts Institute of Technology (1)
- National Center for Biotechnology Information - NCBI (12)
- Publishing Network for Geoscientific & Environmental Data (294)
- QSpace: Queen's University - Canada (3)
- QUB Research Portal - Research Directory and Institutional Repository for Queen's University Belfast (1)
- Repositório Científico da Universidade de Évora - Portugal (3)
- Repositório Científico do Instituto Politécnico de Lisboa - Portugal (4)
- Repositório da Produção Científica e Intelectual da Unicamp (38)
- Repositório digital da Fundação Getúlio Vargas - FGV (1)
- Repositório Institucional da Universidade de Aveiro - Portugal (2)
- Repositório Institucional da Universidade Federal do Rio Grande do Norte (1)
- Repositório Institucional UNESP - Universidade Estadual Paulista "Julio de Mesquita Filho" (37)
- RUN (Repositório da Universidade Nova de Lisboa) - FCT (Faculdade de Cienecias e Technologia), Universidade Nova de Lisboa (UNL), Portugal (7)
- Scielo Saúde Pública - SP (18)
- Scientific Open-access Literature Archive and Repository (1)
- Scottish Institute for Research in Economics (SIRE) (SIRE), United Kingdom (1)
- Universidad de Alicante (4)
- Universidad Politécnica de Madrid (9)
- Universidade Complutense de Madrid (2)
- Universidade do Minho (2)
- Universidade Federal do Pará (1)
- Universidade Federal do Rio Grande do Norte (UFRN) (2)
- Universidade Técnica de Lisboa (2)
- Universita di Parma (1)
- Universitätsbibliothek Kassel, Universität Kassel, Germany (3)
- Université de Lausanne, Switzerland (39)
- Université de Montréal, Canada (4)
- University of Michigan (2)
- University of Queensland eSpace - Australia (97)
- University of Washington (5)
- Worcester Research and Publications - Worcester Research and Publications - UK (1)
Resumo:
Active machine learning algorithms are used when large numbers of unlabeled examples are available and getting labels for them is costly (e.g. requiring consulting a human expert). Many conventional active learning algorithms focus on refining the decision boundary, at the expense of exploring new regions that the current hypothesis misclassifies. We propose a new active learning algorithm that balances such exploration with refining of the decision boundary by dynamically adjusting the probability to explore at each step. Our experimental results demonstrate improved performance on data sets that require extensive exploration while remaining competitive on data sets that do not. Our algorithm also shows significant tolerance of noise.