Open zhangxingsky opened 1 year ago
Unfortunately I don't have the answer to this since the initial preprocessing (and splitting) of these datasets from the raw UCI data was not done by myself but rather by the authors of this paper. I realize that this can make it hard to determine which feature is which if you really want to dig in; however, the datasets are still useful for benchmarking against the prior works using these same splits.
Hello, can you tell me the pre-processing steps for these datasets, I find it a bit different from the original UCI dataset