Figure 4

Classifier design tab. The training data are imported into the "DATA SETS" frame. The "Imported files" can be assigned to either "Training data files" or to "Testing data files" by clicking on the respective buttons. The "CLASSES" frame allows selecting and combining cases to be used for the classifier as training and to establish their name and composition. On "Class name", one can write down the name of the desired class. "Tumour types (number of cases)" displays the number of cases of each type in the training dataset, which can be assigned to the preferred class for classification. Several types already set in the "Training data files" can be merged into the same classification class, therefore allowing different combinations of training data types, for hypothesis testing. The "FEATURE SELECTION AND EXTRACTION" frame allows choosing the desired feature selection or extraction technique and the evaluation method. In this example the "Sequential Forward FS" and "Correlation-based Feature Subset Selection" have been chosen. Clicking on the "Run Feature Selection or Extraction" button below gives the resulting features. "DS1" means "Dataset one", since it is possible to concatenate two spectra from the same case obtained under different acquisition conditions and therefore the first one entered would be DS1. The "CLASSIFIER" frame allows the user to choose the spectral range (in ppm) which will be the desired region of interest for feature selection or extraction and for classification. The "Run classifier" button allows starting the classification with the selected "Classification method" (currently, Fisher LDA).