图 书 馆
主动学习生物医学引文筛选Active Learning for Biomedical Citation Screening |
|
| 课程网址: | http://videolectures.net/kdd2010_wallace_albc/ |
| 主讲教师: | Byron C. Wallace |
| 开课单位: | 布朗大学 |
| 开课时间: | 2010-10-01 |
| 课程语种: | 英语 |
| 中文简介: |  |
| 课程简介: | Active learning (AL) is an increasingly popular strategy for mitigating the amount of labeled data required to train classifi ers, thereby reducing annotator e ffort. We describe a real-world, deployed application of AL to the problem of biomedical citation screening for systematic reviews at the Tufts Evidence-based Practice Center. We propose a novel active learning strategy that exploits a priori domain knowledge provided by the expert (speci fically, labeled features) and extend this model via a Linear Programming algorithm for situations where the expert can provide ranked labeled features. Our methods outperform existing AL strategies on three real-world systematic review datasets. We argue that evaluation must be specifi c to the scenario under consideration. To this end, we propose a new evaluation framework for fi nite-pool scenarios, wherein the primary aim is to label a fixed set of examples rather than to simply induce a good predictive model. We use a method from medical decision theory for eliciting the relative costs of false positives and false negatives from the domain expert, constructing a utility measure of classi fication performance that integrates the expert preferences. Our fi ndings suggest that the expert can, and should, provide more information than instance labels alone. In addition to achieving strong empirical results on the citation screening problem, this work outlines many important steps for moving away from simulated active learning and toward deploying AL for real-world applications. |
| 关 键 词: | 主动学习; 生物医学引文; 医疗决策理论 |
| 课程来源: | 视频讲座网 |
| 最后编审: | 2020-06-06：魏雪琼(课程编辑志愿者) |
| 阅读次数: | 72 |