Rdimtools (version 1.0.4)

do.rsr: Regularized Self-Representation

Description

Given a data matrix $X$ where observations are stacked in a row-wise manner, Regularized Self-Representation (RSR) aims at finding a solution to following optimization problem $$\text{min}\Vert X-XW{\Vert }_{2,1}+\lambda \Vert W{\Vert }_{2,1}$$ where $\Vert W{\Vert }_{2,1}=\sum _{i=1}^m\Vert W_{i:}{\Vert }_2$ is an ${\ell }_{2,1}$ norm that imposes row-wise sparsity constraint.

Usage

do.rsr(
  X,
  ndim = 2,
  preprocess = c("null", "center", "scale", "cscale", "whiten", "decorrelate"),
  lbd = 1
)

Arguments

X

an $(n\times p)$ matrix or data frame whose rows are observations and columns represent independent variables.

ndim

an integer-valued target dimension.

preprocess

an additional option for preprocessing the data. Default is "null". See also aux.preprocess for more details.

lbd

nonnegative number to control the degree of self-representation by imposing row-sparsity.

Value

a named list containing

Y

an $(n\times ndim)$ matrix whose rows are embedded observations.

featidx

a length-$ndim$ vector of indices with highest scores.

trfinfo

a list containing information for out-of-sample prediction.

projection

a $(p\times ndim)$ whose columns are basis for projection.

References

zhu_unsupervised_2015Rdimtools

Examples

# NOT RUN {
## load iris data
data(iris)
set.seed(100)
subid = sample(1:150,50)
X     = as.matrix(iris[subid,1:4])
label = as.factor(iris[subid,5])

#### try different lbd combinations
out1 = do.rsr(X, lbd=0.1)
out2 = do.rsr(X, lbd=1)
out3 = do.rsr(X, lbd=10)

#### visualize
opar <- par(no.readonly=TRUE)
par(mfrow=c(1,3))
plot(out1$Y, pch=19, col=label, main="RSR::lbd=0.1")
plot(out2$Y, pch=19, col=label, main="RSR::lbd=1")
plot(out3$Y, pch=19, col=label, main="RSR::lbd=10")
par(opar)
# }
# NOT RUN {
# }