voomDDA

[…] vector machines) and rf (random forests) algorithms are also considered due to their good performances in microarray based classification studies. implementation details of each algorithm, including voomdda classifiers are given in below:, plda1: the data are normalized using the deseq median ratio method. normalized count values are taken as input to the plda algorithm. a five-fold cross […]

[…] data are converted into a continuous scale using this mean and variance relationship. since plda1 and nblda are count-based classifiers, the transformations are not applied for these classifiers. voomdda classifiers use the voom method inside the algorithm for transformation. a power transformation is applied for plda2 classifier. the rlog transformation is performed for other classifiers, […]