DDL: Deep Dictionary Learning for Predictive Phenotyping.

DDL: Deep Dictionary Learning for Predictive Phenotyping.
复制标题

DOI:
10.24963/ijcai.2019/812
复制
发表时间:
2019-08
期刊:
IJCAI : proceedings of the conference
影响因子:
--
通讯作者:
Sun J
Sun J
中科院分区:
其他
文献类型:
--
作者:
Fu T;Hoang TN;Xiao C;Sun J

文献摘要

参考文献

被引文献

相似文献

预测性表型是指根据纵向电子健康记录(EHR)数据准确预测下一次临床就诊时将出现的表型。虽然深度学习(DL)模型最近在预测性表型方面表现出很强的性能,但它们需要访问大量的标记数据,而这些数据的获取成本很高。为了解决标签不足的挑战,我们提出了一个用于表型识别的深度词典学习框架(DDL),它利用未标记的数据作为补充信息源来生成更好、更简洁的数据表示。我们在多个EHR数据集上的经验评估表明,在需要患者表型分析的各种临床任务中,DDL的表现优于现有的预测性表型分析方法。结果还表明,与只使用标记数据的现有方法相比,未标记数据可以用来生成更好的数据表示,从而帮助提高DDL的表型性能。
Predictive phenotyping is about accurately predicting what phenotypes will occur in the next clinical visit based on longitudinal Electronic Health Record (EHR) data. While deep learning (DL) models have recently demonstrated strong performance in predictive phenotyping, they require access to a large amount of labeled data, which are expensive to acquire. To address this label-insufficient challenge, we propose a deep dictionary learning framework (DDL) for phenotyping, which utilizes unlabeled data as a complementary source of information to generate a better, more succinct data representation. Our empirical evaluations on multiple EHR datasets demonstrated that DDL outperforms the existing predictive phenotyping methods on a wide variety of clinical tasks that require patient phenotyping. The results also show that unlabeled data can be used to generate better data representation that helps improve DDL’s phenotyping performance over existing methods that only uses labeled data.
DOI: 10.1145/3097983.3098126
发表时间: 2017-08
期刊: KDD : proceedings. International Conference on Knowledge Discovery & Data Mining
影响因子: --
作者:
Choi E;Bahadori MT;Song L;Stewart WF;Sun J
通讯作者: Sun J
DOI: 10.1016/j.jbi.2016.10.007
发表时间: 2016-12-01
影响因子: 4.5
作者:
Beaulieu-Jones, Brett K.;Greene, Casey S.
通讯作者: Greene, Casey S.
DOI: 10.1109/access.2016.2611583
发表时间: 2016-01-01
期刊: IEEE ACCESS
影响因子: 3.9
作者:
Tariyal, Snigdha;Majumdar, Angshul;Vatsa, Mayank
通讯作者: Vatsa, Mayank