13040_2019_193_MOESM3_ESM.pdf (419.13 kB)
Additional file 3: of Predicting opioid dependence from electronic health records with machine learning
journal contribution
posted on 2019-01-29, 05:00 authored by Randall Ellis, Zichen Wang, Nicholas Genes, Avi MaâayanFigure S3. Receiver operating characteristic curves, normalized, and non-normalized confusion matrices for classifiers using all features (A, B), the top 10 by Gini importance (C, D), and the top 10 by p-value from the Wilcoxon rank-sum test (E, F). (PDF 419 kb)