Possible prediction types are: Classification: Labels or class probabilities (including labels). Regression: Numeric or response or standard errors (including numeric response). Survival: Linear predictor or survival probability.
For complex wrappers the predict type is usually also passed down the encapsulated learner in a recursive fashion.
Arguments
- learner
(Learner |
character(1))
The learner. If you pass a string the learner will be created via makeLearner.- predict.type
(
character(1))
Classification: “response” or “prob”. Regression: “response” or “se”. Survival: “response” (linear predictor) or “prob”. Clustering: “response” or “prob”. Default is “response”.
See also
Other predict:
asROCRPrediction(),
getPredictionProbabilities(),
getPredictionResponse(),
getPredictionTaskDesc(),
predict.WrappedModel(),
setPredictThreshold()
Other learner:
LearnerProperties,
getClassWeightParam(),
getHyperPars(),
getLearnerId(),
getLearnerNote(),
getLearnerPackages(),
getLearnerParVals(),
getLearnerParamSet(),
getLearnerPredictType(),
getLearnerShortName(),
getLearnerType(),
getParamSet(),
helpLearnerParam(),
helpLearner(),
makeLearners(),
makeLearner(),
removeHyperPars(),
setHyperPars(),
setId(),
setLearnerId(),
setPredictThreshold()