Hi,
I was wondering if we are allowed to use the test data for training unsupervised or semi-supervised models? I understand we will be receiving the validation labels, so it’s clear we will be able to use the validation data for training, but it’s not clear if the test data can be used in training.
Thanks,
Jim
Hi @jimking100, thank you for checking.
Competitors are not allowed to use the test set data for unsupervised or semi-supervised training. This is covered in the competition rules in the “Entry submissions” section, excerpted below:
Unless otherwise specified on the Competition Website, for the purposes of quantitative evaluation of Submissions, Participants agree to process each test data sample independently without the use of information from other cases in the test set. By default, this precludes using information gathered across multiple test samples during training, for instance through pseudo labeling. Eligible Submissions and models must be able to run inference on new test data automatically, without retraining the model.