Hybrid Control Policy for Artificial Pancreas via Ensemble Deep Reinforcement Learning

Abstract

The study presents novel control policies for AP systems, affirming their great potential for efficient closed-loop glucose control.