Machine Learning for Prediction in Electronic Health Data

Machine learning for prediction in electronic health data has been deployed for many clinical questions during the last decade. Machine learning methods may excel at finding new features or nonlinear relationships in the data, as well as handling settings with more predictor variables than observations. However, the usefulness of both these data and machine learning has varied. Electronic health data often have quality issues (eg, missingness, misclassification, measurement error), and machine learning may perform similarly to standard techniques for some research questions. Ensembles (running multiple algorithms and either selecting the single best algorithm or creating a weighted average) can help mitigate the latter concern.