My research group is publishing an early sepsis prediction research using the eICU data. According to the journal policy, providing open access to data and codes is recommended.
We understand that we cannot provide access to the original data. However, in our research , we first extracted features from the time series of patient physiological and lab data using multiple time windows, such as mean, maximum, minimum, skewness and slope. We also compute the probabilities of sepsis using machine learning models.
Are we allowed to share the processed data such as the feature data, or the probabilities? If so, it would help us greatly to provide repeatable expriments for the readers. We will make sure no individual can be identified from the shared data.
Hello,
My research group is publishing an early sepsis prediction research using the eICU data. According to the journal policy, providing open access to data and codes is recommended.
We understand that we cannot provide access to the original data. However, in our research , we first extracted features from the time series of patient physiological and lab data using multiple time windows, such as mean, maximum, minimum, skewness and slope. We also compute the probabilities of sepsis using machine learning models.
Are we allowed to share the processed data such as the feature data, or the probabilities? If so, it would help us greatly to provide repeatable expriments for the readers. We will make sure no individual can be identified from the shared data.
Thanks!