Active learning for clinical text classification: is it better than random sampling?
Open Access
- 1 September 2012
- journal article
- research article
- Published by Oxford University Press (OUP) in Journal of the American Medical Informatics Association
- Vol. 19 (5) , 809-816
Abstract
Objective This study explores active learning algorithms as a way to reduce the requirements for large training sets in medical text classification tKeywords
This publication has 31 references indexed in Scilit:
- Predicting sample size required for classification performanceBMC Medical Informatics and Decision Making, 2012
- Adoption of telemedicine: from pilot stage to routine deliveryBMC Medical Informatics and Decision Making, 2012
- Accelerating the annotation of sparse named entities by dynamic sentence selectionBMC Bioinformatics, 2008
- Mayo Clinic NLP System for Patient Smoking Status IdentificationJournal of the American Medical Informatics Association, 2008
- Heuristic Sample Selection to Minimize Reference Standard Training Set for a Part-Of-Speech TaggerJournal of the American Medical Informatics Association, 2007
- Using active learning to annotate microscope images of parasite eggsArtificial Intelligence Review, 2006
- Active Learning with Support Vector Machine Applied to Gene Expression Data for Cancer ClassificationJournal of Chemical Information and Computer Sciences, 2004
- Active Learning with Support Vector Machines in the Drug Discovery ProcessJournal of Chemical Information and Computer Sciences, 2003
- Two biomedical sublanguages: a description based on the theories of Zellig HarrisJournal of Biomedical Informatics, 2002
- 10.1162/153244302760185243Applied Physics Letters, 2000