Identifying determinants of persistent MRSA bacteremia using mathematical modeling
Fig 5
Key determinants to distinguish resolving and persistent bacteremia.
(A) Pearson correlation coefficients were calculated between each feature and the binary response variable (0: RB, 1: PB). The absolute values of the coefficients are shown. (B) Weights calculated by QPFS methods are shown. (C) The sole ranked feature by QPFS, cP, was used to build a logistic regression model. For the comparison, all parameter and all parameter except for cP were applied to the classification model. Best accuracy, ROC curve, and area under ROC curve are graphed for each model. (D) Two-dimensional scatter plot for cP and swN2P, which showed 2nd high correlation with target response (Fig5 A) with probability density for RB and PB.