Validation Data

[tmt]

For the validation data, do we need to choose and select from the training dataset provided in the challenge website in this link: https://physionet.org/content/i-care/1.0/#files, and split it into training and validation dataset, or is there a separate link which provides the validation dataset itself?

[PhysioNet Challenge]

Hello,

We share the training data that we use to train your model, but we do not share the validation and test data that we use to evaluate your model. This allows us to evaluate the models in a consistent and principled way.

To debug your code, you will need to (1) remove the labels and (2) create 12, 24, 48, and 72 hour versions of the data that you are using with your trained model, i.e., the run_model script. We will use the steps described here to prepare the validation and test data, and you can perform the same steps to test your code:
https://physionetchallenges.org/2023/submissions/#faq

To develop and evaluate your model, we encourage you to perform cross-validation on the training data.

Best,
Matt
(On behalf of the Challenge team.)

Please post questions and comments in the forum. However, if your question reveals information about your entry, then please email info at physionetchallenge.org. We may post parts of our reply publicly if we feel that all Challengers should benefit from it. We will not answer emails about the Challenge to any other address. This email is maintained by a group. Please do not email us individually.