CLASSIF1 Data Pattern Classification

< fig.1    > fig.2

- The parameter values of the data columns selected by data sieving are transformed into -, 0, or + depending of their position below the respective lower, between the lower and upper or above the upper percentiles and classified as a data pattern.
- This provides better results (fig.6A/6B) than classifying according to any one of the ROC (receiver operating characteristic) curves above.

< CLASSIF1    > fig.2