Dear organizers,

85 views
Skip to first unread message

Nina

unread,
Jul 19, 2021, 8:38:41 AM7/19/21
to physionet-challenges
In the output csv files, do we need to provide prediction for all classes (including those not used for scoring) or we can simply ignore those classes? In test , we will predict zero for all classes if the recording doesn't belong to any of scored classes (yes we will scarify performance but this will help us speed up training and testing).

Regards,

PhysioNet Challenge

unread,
Jul 19, 2021, 8:39:26 AM7/19/21
to physionet-challenges
Dear Nina,

You do not need to provide outputs for all of the classes or diagnoses:
https://physionetchallenges.org/2021/#algorithms

The scoring code assumes that missing outputs for scored classes are zero. In other words, if you do not provide an output for a scored class, then the scoring code will treat it as if your model decided that the diagnosis was not present in the recording.

The scoring code ignores unscored classes, so it does not matter whether or not you provide outputs for them.

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

https://PhysioNetChallenges.org/
https://PhysioNet.org/

Please post questions and comments in the forum. However, if your question reveals information about your entry, then please email challenge at physionet.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.
Reply all
Reply to author
Forward
0 new messages