Abstract

Blood glucose level prediction (BGLP) is a challenging task for AI researchers,
with the potential to improve the health and well-being of people with
diabetes. For this task, a set of tools for end-to-end time series ML project
lifecycle has been developed in a secured Kubernetes environment. It includes
data ingestion, storage, visualization, and data and ML pipelines. The work
builds on a baseline set by BGLP challenge solutions on a reference OhioT1DM
dataset applies the algorithms to a private dataset provided by the First
Faculty of Medicine of Charles University, and aims to explore new methods to
improve predictive accuracy and explainability.

Slides

Lecture Recordings

Readings

1. https://sites.google.com/view/kdh-2020/bglp-challenge 
   
2. http://smarthealth.cs.ohio.edu/bglp/bglp-results.html

Catering

Hrozno