Name
Title | ||
|---|---|---|
Estimating summary statistics for electronic health record laboratory data for use in high-throughput phenotyping algorithms. |
Abstract | ||
|---|---|---|
•The PopKLD algorithm selects a parametric model that faithfully summarizes the laboratory data.•The PopKLD algorithm selects a parametric model that preserves known physiologic relationships.•The PopKLD summary has more discernable information and an empirical mean or variance.•The PopKLD algorithm is automatable and its output can be the input for phenotyping algorithms.•The PopKLD summary can be discretized for machine learning algorithms with categorical input. |
Year | DOI | Venue |
|---|---|---|
2018 | 10.1016/j.jbi.2018.01.004 | Journal of Biomedical Informatics |
Keywords | Field | DocType |
Electronic health record,Kullback-Leibler divergence,Summary statistic,phenotyping,Laboratory tests | Health care,Data mining,Parametric model,Computer science,Categorical variable,Model selection,Algorithm,Topic model,Principle of maximum entropy,Standard deviation,Intensive care | Journal |
Volume | ISSN | Citations |
78 | 1532-0464 | 1 |
PageRank | References | Authors |
0.35 | 16 | 6 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| David J Albers | 1 | 30 | 3.90 |
| Noémie Elhadad | 2 | 6 | 1.22 |
| Jan Claassen | 3 | 10 | 2.65 |
| R. Perotte | 4 | 1 | 0.69 |
| Andrew Goldstein | 5 | 1 | 0.69 |
| George Hripcsak | 6 | 24 | 11.67 |