about dataset

[wzz-z]

Hi author, thanks for your contribution, I am trying to use your dataset for training, but in doing so I have found that the Bicycle_01 folder in the ED24 directory, for example, seems to contain 41 levels of noise events from the noise labels in the fifth column of the csv, corresponding to the 41 csv files in that directory. Then, with only 100 samples available for training at the same noise level, is the sample size too sparse?