Hello @EHRChallengeParticipants , We are happy to announce that the new synthetic data is available for download. Check out the [data folder](syn20685954) for data descriptions and terms of use. We are making two new synthetic datasets available: the Fast Lane Synthetic data and the Full Synthetic data. Both of these datasets have been adapted from the Synpuf dataset that was previously used to more accurately reflect the UW Challenge Data. We have made changes so that the new synthetic data has the same columns, similar variable types (ex. adding float values to value_as_number in the measurement table), and similar record distribution (same percentage of patients with 1 measurement record, 2 measurement records, etc.) as the UW Challenge Data. Per previous requests, we are also including a death table in all the training and evaluation datasets. The only difference between the two new datasets is the size. The Fast Lane Synthetic data is approximately 10% the size of the UW Challenge data and the Full Synthetic data is 100% the size of the UW Challenge data. The Fast Lane synthetic data will be used for both the Fast Lane Queue and the Main Challenge Submission Queue to confirm valid model submissions. At this point in time, the Fast Lane Synthetic data is available for download. The Full Synthetic data is being uploaded (approx 14GB zipped) and should appear in the next several hours in the linked data folder. Thank you and if you have questions feel free to leave them in this discussion thread, Thank you, Tim

Created by Timothy Bergquist trberg

New Synthetic Data Available page is loading…