crda.classify: Classification Example using Compressive Regularized...

Description Usage Arguments Details Value Author(s) See Also Examples

Description

The function crda.cv performs classification using CRDA-variants for a partially synthetic dataset.

Usage

1
2
crda.classify(L = 10, nK = 10, q = Inf, prior = "uniform",
  centerX = FALSE)

Arguments

L

Number of runs, i.e., training and test splits.

nK

Number of candidates in 5-fold CV-grid for finding joint-sparsity level.

q

Type of Lq,1 norm, default is Linf-norm.

prior

Type of prior class probabilities, either 'uniform' (default) or 'estimated'.

centerX

Flag for grand-mean centering of test dataset using grand-mean of training dataset.

Details

crda.classify

Value

An object res of class crda.classify with the following attributes:

funCall

The call to the crda.classify function.

ACT

Average computational time (ACT) in seconds over L-runs.

AveTER

Average test error rate (TER) over L-runs.

AveFSR

Average feature selection rate (FSR) over L-runs.

AveFPR

Average false positive rate (FPR) over L-runs.

AveFNR

Average false negative rate (FNR) over L-runs.

Author(s)

Muhammad Naveed Tabassum and Esa Ollila, 2018

See Also

crda, crda.setup3, crda.regparam, crda.cv

Examples

1
2
3
4
5
crda.classify()
crda.classify(L = 1)
crda.classify(L = 1, nK = 5)
crda.classify(L = 1, prior = 'estimated')
crda.classify(L = 1, centerX = TRUE)

mntabassm/compressiveRDA documentation built on May 31, 2019, 5:22 p.m.