Asthma exacerbation prediction

This is the codebase for the paper https://www.medrxiv.org/content/medrxiv/early/2019/11/29/19012161.full.pdf
Asthma Exacerbation Prediction and Interpretation based on Time-sensitive Attentive Neural Network: A Retrospective Cohort Study , which is used to predict the risk of asthma exacerbation among asthma patients according to a history of EHR data. The model can be also scaled to other sequential prediction tasks. Please refer to the paper for more details

Environment

Python 3.0+
Tensorflow 1.14+, GPU
Numpy, Scikit-learn, etc.

Demo script

cd src/
sh run.sh

Sample data format

code: a list (list of visits) of list (list of code in each visit)
time: a list (list of visits) of list (time to the first visit, time to the previous visit, time to the last visit)
You can choose any time you want or choose all
There is also a sample data in the data/ directory (compressed using pickle), you can load it using pkl.load