PredictPerformance: Predicts suitable outlier methods from meta-features of the...
In sevvandi/outselect: Algorithm selection for unsupervised outlier detection

Description Usage Arguments Value Examples

Predicts suitable outlier methods from meta-features of the dataset using trained random forest models.

1	PredictPerformance(ftrs, models)

`ftrs`	Meta-features of the dataset.
`models`	The trained random forest models. These are obtained from the function `TrainModels`.

The prediction probabilities for each of the outlier method. For example, a probability of 0.78 for the first outlier method means the probability that the first method is good for this dataset is 0.78.

## Not run: 
data(Arrhythmia_withoutdupl_05_v05)
dat <- Arrhythmia_withoutdupl_05_v05
feat <- ComputeMetaFeaturesMM(dat)
fit <- TrainModels(1,1,1)
out <- PredictPerformance(feat, fit)

## End(Not run)