Expert Guided Natural Language Processing Using One-Class Classification
Conclusions 1C-SVMs trained on expert-selected relevant text sections perform better than 2C-SVMs classifiers trained on either snippets or whole notes when applied to realistically imbalanced data with low prevalence of the positive class.
Source: Journal of the American Medical Informatics Association - Category: Information Technology Authors: Joffe, E., Pettigrew, E. J., Herskovic, J. R., Bearden, C. F., Bernstam, E. V. Tags: Research and Applications Source Type: research