Importing 1D conv Net, LSTM

[kelleyjbrady]

Hi in your paper the authors state that Extensive experimentation was performed to determine the optimal methods to format the chronological data as input and construct and train the most accurate deep learning model for outpatient forecasting within this data set. Would you be willing to share any more details of the 'negative' results? Based on your imports, I'm guessing for example you tried using 1D convnets and LSTM's, but then settled on the GRU architecture because of better performance?