Training & Test Data


Training Data is the data that MultiRate uses when running the analysis and calculating factors. Test data is a separate set of data used to evaluate the factors calculated by the training data. Using test data is optional, but highly recommended as it is useful to see how the model performs with data that was not used in calculating the factors in the first place.


Training and test data can be selected separately. If they come from the same table, a Filter should be used to divide the data (adding a Random Characteristic can be useful for this). If the training data and test data are located in separate tables, the field names and data types for both tables must be identical.