Hello, I'm trying to reproduce the results of a paper (objective is to use NMF and PCA to find clusters in data) but I'm no math expert and am having trouble understanding what they did in here:
"Two hundred iterations of fivefold resampling, training on a partition of approximately 80% of the samples, were performed to achieve stable NMF results. For each of these 200 data partitions NMF was performed"
I think that they got 80% of the samples to use as a training model and did this 200 times? but the rest is a bit confusing. I don't understand what is this fivefold resampling. When I think of resampling I think of bootstrap but I'm not sure they did this here? Does this fivefold means k-fold cross validation?
Any help is appreciated. Thanks in advance