Prioritization of processes based on user persona and reinforcement learning

Abstract

Settings on an information handling system may be adjusted to set priority levels for processes executing on the information handling system in view of desired operational characteristics of the information handling system for a user persona and in view of expected future events for the information handling system. A score may be generated based on a user persona (e.g., whether a user is a light gamer, heavy gamer, corridor warrior, or desk worker) and expected future computer contexts (e.g., an expectation that a user will play a game in one hour). That score may be used to determine policies (e.g., high performance, balanced, or battery saver) to implement through settings on the computer system. Consideration of user persona classifications, associated group behaviors, and dynamic system contexts (including resource extremas, location, temporal context, and predicted future events) improve use of system resources through prioritization and governing of diverse optimization methods.