Figure 5: Logistic regression classification results using primary and augmented predictor models. (a) Primary predictor model (PPM): the identification procedure incorporated , “radiologic finding,” , and “abnormal radiologic density, nodular” in the first two steps. The process stopped at step 2 as new predictor cannot make any improvement in classification. (b) Augmented predictor model (APM): two new features, and , were included in the model in the first two steps of the procedure. The predictors, and , represent the squared sums of ontological features and their absolute values, respectively. The identification proceeds to step 5 that extends the feature space into three predictor dimensions, , , and , representing “radiologic finding,” “mass of body region,” and “imaging result abnormal.”