This is the model "Heterogeneous Event LSTM"(HELSTM) for the paper "Learning the Joint Representation of Heterogeneous Temporal Events for Clinical Endpoint Prediction"

You can find the paper here

task.py shows how we used HELSTM to do end-to-end prediction. Data will be load from "/data/", or you can change data_path in task.py.

We divided all data into three parts: train_data, valid_data and test_datam, and processed data into h5py format, containing "time", "label", "event", "feature_id" and "feature_value". Data sequences should have the same length (by padding 0 in the end) and we set length to 1000 in task.py.

How to use the data generator:

1.put extractor.cpp and dataExt.py under the same directory as the origin mimic data.

2.compile and run extractor.cpp as: g++ extractor.cpp -o extractor ./extractor

3.run dataExt.py as python dataExt.py