I'm aware of and have worked with many datasets in Classical ML as well as DL. I am also aware of some of the standard datasets in DL (for example ImageNet for Image Classification, etc.)

However, I was wondering if there are any standard datasets (or benchmarks for accuracy) for the Classical methods such as Regression, GBM, SVM, etc. More specifically, are there any standard datasets that can be used to measure the accuracy of a new method?

Given that most of the Classical methods are very old, the datasets they would've used to test their methods may not be relevant today.

If there are no such standards, can you comment on the class of applications you would like to see if someone were to create their own standard dataset?



1 Answer 1


Before the deep learning wave, the the UCI dataset repository was widely used.

It contains classic (and rather small) datasets that were very relevant in the old days, like the Iris dataset for classification.

In each dataset page, you can find papers citing the dataset.

  • $\begingroup$ Thanks. That helped. $\endgroup$ Jul 10, 2021 at 16:34
  • $\begingroup$ As you stated, the UCI dataset was relevant in old days. I actually needed the datasets relevant today. I need some huge datasets, but which are also standard. Thanks for the answer, btw $\endgroup$ Jul 10, 2021 at 17:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.