CART Class Prediction: Classify Unknown Sample

To evaluate an existing Classification And Regression Trees (CART) classifier, run the CART module with the classifier (model) file and a data set containing the unknown samples.

Before you begin

Classifying an unknown sample requires:

A *.cart.model file that contains the CART classifier.
Example file: all_aml_train.cart.model.
A gene expression data set containing the unknown samples:
- GCT or RES file that contains gene expression data.
  Example file: all_aml_test.gct.
- CLS file that arbitrarily assigns the class of each sample in the gene expression data to one of the classes in the training set.
  Example file: all_aml_test_classify_CART.cls.
learn more:
file formats

Step 1: CART

To classify unknown samples using a previously built CART classifier, run the CART module. Use the model file parameter to specify the classifier (*.cart.model) and the test filename and test class filename parameters to specify the data set containing the unknown samples. The module creates a prediction results file (*.pred.odf) that lists each sample with its actual class ('unknown') and its predicted class.

learn more:
CART

Step 2: View results

To view the prediction results file (*.pred.odf), use the PredictionResultsViewer module. The viewer lists each sample with its actual class ('unknown') and its predicted class.

Considerations

Ignore the actual class, which was unknown.
Ignore the error rates, which are designed to evaluate the class predictor against known data.

learn more:
PredictionResultsViewer