RaSEn (version 3.0.0)

predict.RaSE: Predict the outcome of new observations based on the estimated RaSE classifier (Tian, Y. and Feng, Y., 2021).

Description

Predict the outcome of new observations based on the estimated RaSE classifier (Tian, Y. and Feng, Y., 2021).

Usage

# S3 method for RaSE
predict(object, newx, type = c("vote", "prob", "raw-vote", "raw-prob"), ...)

Arguments

object

fitted 'RaSE' object using Rase.

newx

a set of new observations. Each row of newx is a new observation.

type

the type of prediction output. Can be 'vote', 'prob', 'raw-vote' or 'raw-prob'. Default = 'vote'.

• vote: output the predicted class (by voting and cut-off) of new observations. Avalilable for all base learner types.

• prob: output the predicted probabilities (posterior probability of each observation to be class 1) of new observations. It is the average probability over all base learners. Avalilable only when base leaner is not equal to 'svm' and 'tree'.

• raw-vote: output the predicted class of new observations for all base learners. It is a n by B1 matrix. n is the test sample size and B1 is the number of base learners used in RaSE. Avalilable for all base learner types.

• raw-prob: output the predicted probabilities (posterior probability of each observation to be class 1) of new observations for all base learners. It is a n by B1 matrix. Avalilable only when base leaner is not equal to 'svm' and 'tree'.

...

additional arguments.

Value

depends on the parameter type. See the list above.

References

Tian, Y. and Feng, Y., 2021. RaSE: Random subspace ensemble classification. Journal of Machine Learning Research, 22(45), pp.1-93.

See Also

Rase.

Examples

# NOT RUN {
set.seed(0, kind = "L'Ecuyer-CMRG")
train.data <- RaModel("classification", 1, n = 100, p = 50)
test.data <- RaModel("classification", 1, n = 100, p = 50)
xtrain <- train.data$x
ytrain <- train.data$y
xtest <- test.data$x
ytest <- test.data$y

model.fit <- Rase(xtrain, ytrain, B1 = 100, B2 = 100, iteration = 0, base = 'lda',
cores = 2, criterion = 'ric', ranking = TRUE)
ypred <- predict(model.fit, xtest)
mean(ypred != ytest)
# }
# NOT RUN {
# }