Code for the experiments run in:
Jeeheh Oh, Jiaxuan Wang, Jenna Wiens. “Learning to Exploit Invariances in Clinical Time-Series Data Using Sequence Transformer Networks.” Machine Learning for Healthcare Conference, 2018.
SeqTN_functions.py
- HP_search()
- test()
- train()
- SeqTN() : Sequence Transformer Networks
- CNN() : the baseline CNN
Data can be downloaded here: https://mimic.physionet.org/
Data can be preprocessed via code found here: https://github.com/YerevaNN/mimic3-benchmarks