Test Data

Dear Organizers, I am writing to inquire about the test data provided in the published dataset on Zenodo. Are the test data included in this dataset identical to those used in the challenge? If not, I would like to inquire whether there will be an opportunity for post-challenge submissions. This would enable participants to compare their models from the challenge with any subsequent data provided.

Thanks in advance! I appreciate your assistance and clarification.

The test data provided at Zenodo (https://zenodo.org/records/8417503) is the same as the test data used for the test phase.

Also, we internally discussed re-opening the test phase submission, but after considering extra cost and sustainability/maintainability, we eventually decided not to open it. Thank you for understanding.

Instead, we provided an evaluation code here: https://github.com/lunit-io/ocelot23algo/tree/main/evaluation. With your prediction in json format, you can compute test performance.