Size training dataset

[Lore Van Santvliet]

Dear Physionet Challenge Organisers,

In the submission instructions, I read the following:

"For training your model on the training set, we impose a 24 hour time limit on a subset (~1000 records) of the training set and a 168 hour limit on the entirety (~22,000 records) of the training set. For running your trained model on the validation set (~1000 records), we impose a 24 hour time limit.

Doesn't the training set include the CODE-15% dataset (300,000 records) + Sami-Trop + PTB-XL? Then how can the entire training set only be ~22,000 records?

Kind regards,
Lore

[PhysioNet Challenge]

Dear Lore,

Good catch. For the unofficial phase, we are currently allowing 72 hours for training on the training set (~300k+ records) and 24 hours for inference on the validation set (~30k records).

This part of the instructions was a relic from last year's Challenge, which used ECG image data, but I have now updated it for this year's Challenge, which uses ECG time series data. It took more time to generate and process the ECG images, which is why we provided longer run times despite the smaller numbers of records last year.

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.