Skip to content

LIG-Doctor - patient trajectory prediction using Deep Learning framework Theano

License

Notifications You must be signed in to change notification settings

jfrjunio/LIG-Doctor

Repository files navigation

LIG-Doctor

#testing the installation
python import theano #should execute smoothly

#after all set
#model training - builds model to be saved in a file called model.npz:
"python ./LIG-Doctor.py ./mimic-90-10_02_271 ./model"

#model testing:
"python ./LIG-Doctor-test.py ./mimic-90-10_02_271 ./model.npz"

These input files are outputs of script preprocess_mimiciii.py. We provide actual input files for reproduction of our results. So, runing preprocess_mimiciii.py is, initially, optional. DISCLAIMER: we built the data file using the mimic-III dataset; but the file is nothing but a bunch of numbers in binary format; in order to make sense out of the file, one must download mimic-III from https://mimic.physionet.org

About

LIG-Doctor - patient trajectory prediction using Deep Learning framework Theano

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages