Skip to content

Used DeepAR to forecast hourly patient census. Part my field session project for Colorado School of Mines in Summer 2020.

Notifications You must be signed in to change notification settings

cgraflund/DeepAR-Patient-Forecasting

Repository files navigation

DeepAR Patient Forecasting

This project was a part of the Colorado School of Mines Summer 2020 Field Session for Medecipher.

The goal of this project was to experiment and find a forecasting model that had better error metrics than Medecipher's current model, Arimax.

Amazon's machine learning algorithm, DeepAR, was used for this notebook.

For privacy reasons, the patient dataset is not included in this repository.

About

Used DeepAR to forecast hourly patient census. Part my field session project for Colorado School of Mines in Summer 2020.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published