Skip to content

Classification & Feature Selection for predicting blood glucose variability in Type 2 Diabetes

License

Notifications You must be signed in to change notification settings

christophvoe/Avoid_BG_fluctuation_TD2

Repository files navigation

DATA ARCHIVE - Avoid_BG_fluctuation_TD2

This repository shows a way to assess the risk of fluctuating blood glucose levels in Diabetes Type 2 patients. For this repository synthetic data is used as existing patient data is highly sensible and must not be publicly available. For now simple classification algorithms are used which do not yield a satisfactory result yet.


For reproducing the workflow please run all the corresponding scripts. The file Synthetic_Data can't be reproduced as the existing patient data is not provided. To see the code in action follow the link to the html document.


Files/Folders Description
Simple Classifiers Hyper R-Markdown-script to run classifiers
Simple Classifiers Hypo R-Markdown-script to run classifiers
/Data Folder containing several folders with different data files. Among these Folders: Hypo data, Hyper data, synthetic data
/Functions Folder containing helper functions for classification

For any help with the files in this archive, please contact Christoph Völtzke ([email protected]).


License

The repository is licensed under the apache-2.0 license.

About

Classification & Feature Selection for predicting blood glucose variability in Type 2 Diabetes

Topics

Resources

License

Stars

Watchers

Forks