Skip to main content

< Back to Article

A novel interpretable machine learning system to generate clinical risk scores: An application for predicting early mortality or unplanned readmission in a retrospective cohort study

Fig 3

ShapleyVIC values from 350 nearly optimal logistic regression models, arranged in descending order of average ShapleyVIC values.

Ventilation had zero ShapleyVIC value in all 350 models analyzed and was therefore not plotted. Number of ED visit is within 6 months before current inpatient stay. Number of surgery, ICU stay and HD stay are within 1 year before current inpatient stay. CHF: Congestive heart failure; DBP: diastolic blood pressure; ED: emergency department; HD: high dependency ward; ICU: intensive care unit; LOS: length of stay; MI: Myocardial infarction; PVD: Peripheral vascular disease; PUD: Peptic ulcer disease; SBP: systolic blood pressure; SpO2: blood oxygen saturation.

Fig 3