The example data in the data/ are randomly generated data for the demonstration of the algorithm.
Two types of data is requied for model training and prediction:
gs.file
:.txt
file two columns. The first column if file name index. The second column is the gold standard (0/1), representing the final outbreak of sepsis
0.psv,1
1.psv,1
2.psv,0
3.psv,1
4.psv,0
5.psv,1
6.psv,1
*.psv
:.psv
table files separated by|
, which is the time-series feature records. The header of psv file are the feature names. To note, the first column is the time index.
HR feature_1 featuyre_2 ... feature_n-1 feature_n
0.0 1 0.0 ... 1.3 0.0
1.0 NaN 0.0 ... 0.0 0.0
3.5 NaN 2.3 ... 0.0 0.0
python main.py -g [GS_FILE_PATH] -t [LAST_N_RECORDS] -f [EXTRA_FEATURES]
GS_FILE_PATH
: the path to gold-standards and file path file;LAST_N_RECORDS
: last n records used for prediction. defulat: 16;EXTRA_FEATURES
: addtional features used for prediction. defualt: all features we used in DII Data challenge.
This will generate models, which will be saved under a new directory ./models
This method can be generalized to be used on other hospitalization data. One application of this method is the COVID-19 DREAM Challenge, where this method also achieves top performance.
- For citation, please refer to our latest iScience paper: Assessment of the timeliness and robustness for predicting adult sepsis.
- For protocol(TBD)