Cross-Conformal (CV+) Classification

Constructs conformal prediction sets using K-fold cross-conformal inference.

Usage

cv.conformal.classification(
  fdataobj,
  y,
  newdata,
  covariates.train = NULL,
  covariates.test = NULL,
  ncomp = 3,
  classifier = "lda",
  k.nn = 5,
  score.type = c("lac", "aps"),
  n.folds = 5,
  alpha = 0.1,
  seed = NULL
)

Arguments

fdataobj

An object of class 'fdata' (training data).

y

Class labels (integer vector, 0-indexed).

newdata

An object of class 'fdata' (test data).

covariates.train

Optional scalar covariates for training.

covariates.test

Optional scalar covariates for test.

ncomp

Number of FPC components (default 3).

classifier

Classifier: "lda", "qda", or "knn" (default "lda").

k.nn

k for kNN classifier (default 5).

score.type

Nonconformity score: "lac" or "aps" (default "lac").

n.folds

Number of folds (default 5).

alpha

Miscoverage level (default 0.1).

seed

Random seed.

Value

A list with components:

predicted.classes

Point predictions

set.sizes

Size of each prediction set

average.set.size

Mean prediction set size

coverage

Empirical coverage

Examples

# \donttest{
fd <- fdata(matrix(rnorm(500), 50, 10), argvals = seq(0, 1, length.out = 10))
y <- sample(0:1, 50, replace = TRUE)
cp <- cv.conformal.classification(fd[1:40, ], y[1:40], fd[41:50, ])
# }