Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training an action selection neural network used to select actions to be performed by an agent interacting with an environment. In one aspect, a system comprises a plurality of actor computing units and a plurality of learner computing units. The actor computing units generate experience tuple trajectories that are used by the learner computing units to update learner action selection neural network parameters using a reinforcement learning technique. The reinforcement learning technique may be an off-policy actor critic reinforcement learning technique.

Abstract: A system and method is disclosed for a remote multi-purpose user interface for medical devices and systems within a healthcare/medication delivery system and/or medication information technology system. The multi-purpose user interface has a housing, a processor, a memory, a communications interface for providing communication between the user interface and a medical device/controller and for providing communications between the user interface and a first central computer, and a display for displaying a medical prompt and for displaying medical information received from the first central computer. A system and method is also disclosed for medical data tracking, analyzing and reporting within a healthcare system. The system can further integrate vital signs and infusion pump monitoring and reporting, and allow for enhanced provision of medical care through interface screens which combine this functionality.