Clinical Knowledge Graph Embedding Representation Bridging the Gap between Electronic Health Records and Prediction Models

Matthew Wai Heng Chung; Jianyu Liu; Hegler Tissot

doi:10.1109/ICMLA.2019.00237

Back

Clinical Knowledge Graph Embedding Representation Bridging the Gap between Electronic Health Records and Prediction Models

Conference proceeding

Open access

Clinical Knowledge Graph Embedding Representation Bridging the Gap between Electronic Health Records and Prediction Models

Matthew Wai Heng Chung, Jianyu Liu and Hegler Tissot

2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA), pp 1448-1453

Dec 2019

DOI: https://doi.org/10.1109/ICMLA.2019.00237

Files and links (1)

url

https://discovery.ucl.ac.uk/10094228/1/2019g-ICMLA-Clinical%20Knowledge%20Graph%20Embedding%20Representation.pdfView

SubmittedOpen Access (License Unspecified), Open

Abstract

classification

clustering

Correlation

electronic health records

embedding representation

knowledge graphs

link prediction

Machine learning

Measurement

multi relational data

Pregnancy

Protocols

Task analysis

Training

Learning knowledge embedding representation is an increasingly important technology. However, the choice of hyperparameters is seldom justified and usually relies on exhaustive search. Understanding the effect of hyperparameter combinations on embedding quality is crucial to avoid the inefficient process and enhance practicality of embedding representation along subsequent machine learning applications. This work focuses on translational embedding models for multi-relational categorized data in the clinical domain. We trained and evaluated models with different combinations of hyperparameters on two clinical datasets. We contrasted the results by comparing metric distributions and fitting a random forest regression model. Classifiers were trained to assess embedding representation quality. Finally, clustering was tested as a validation protocol. We observed consistent patterns of hyperparameter preference and identified those that achieved better results respectively. However, results show different patterns regarding link prediction, which is taken as strong evidence that traditional evaluation protocol used for open-domain data does not necessarily lead to the best embedding representation for categorized data.

Metrics

7 Record Views

3 citations in Scopus

See more details

Details

Title: Clinical Knowledge Graph Embedding Representation Bridging the Gap between Electronic Health Records and Prediction Models
Creators: Matthew Wai Heng Chung - University College London
Jianyu Liu - University College London
Hegler Tissot - UCL Australia
Publication Details: 2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA), pp 1448-1453
Publisher: IEEE
Resource Type: Conference proceeding
Language: English
Academic Unit: Information Science (Informatics)
Scopus ID: 2-s2.0-85080920346
Other Identifier: 991021861872704721

Clinical Knowledge Graph Embedding Representation Bridging the Gap between Electronic Health Records and Prediction Models

Files and links (1)

Abstract

Metrics

Details

Drexel University Social media