- CLASSIFICATION (default): The output is the class number.
- REGRESSION: The output is the result of standard regression.
- PROBABILITY: The output is the probability that the classification is correct.
- MULTIPROBABILITY: The output is an array of probabilities that each class is correct ordered by classes seen.
- RAW: The output is an array of the internal representation of the classification process. For example, the raw votes in multi-decision tree models.
- RAW_REGRESSION: The output is an array of the internal representation of the regression process. For example, the raw predictions of multiple regression trees.
Not all classifiers support modes other than CLASSIFICATION.