input:
Dataset:
data to process, each column of the matrix should
contain a data vector
Is data mean-free?:
flag to indicate if data is already mean-free
which results in skipping the mean value calculations and might save
some time
Reduction Mode:
for dimension reduction a subset of principal
components is selected, the mode specifies how this is done
Number of Components:
dimension of data subspace, i.e.,
number of principal components used for dimension-reduction
Variance fraction:
amount of data variance to be represented
in the subspace, e.g., 90% or 95%
output:
Result Dataset: dimension-reduced data set, each column contains a data vector