Analysis of Questionnaire for Traditional Medicine and Development of Decision Support System
Figure 5
Each patient’s prediction probability of test data. Predict with BMI. (a) label is slightly deficiency pattern. (b) label is slightly excess pattern. (a) plotted probability of slightly deficiency pattern test patients’ data (point means each patient) and the closer value of 1 means that the patient is of deficiency pattern and the closer value of 0 means excess pattern. So if each point is above horizon line = 0.5, the point is classified into deficiency pattern. (b) plotted probability of slightly excess pattern test patients’ data (point means each patient) and the closer value of 1 means that the patient is of deficiency pattern and the closer value of 0 means excess pattern. So if each point is below horizon line = 0.5, the point is classified into excess pattern. The patients’ prediction probabilities are higher in Figure 5(a) and lower in Figure 5(b).
