I try to reduce the sample size (inner/outer sample) in a customized 'Th. Ott-SVM', to see what happens with the performance.
At one point I get this message:
What is the minimum sample size?
The error is saying that when the data set gets partioned into a Training and Test set, there are not enough examples to train on. You might need to pass more data rows to it.
The error is saying that when the data set gets partioned into a Training and Test set, there are not enough examples to train on. You might need to pass more data rows to it.