Sequential Optimization of Healthcare Processes using Reinforcement Learning and Sequential Function Models

The optimization of healthcare processes is so important, and it directly affects outcomes. However, traditional methods like Six Sigma and Lean cannot manage the complexities of complex and dynamic healthcare processes. Thus, alternative solutions must strive to enable the optimization of healthcare services. We intend to use reinforcement learning (RL) and sequential function model (SFM) models combined as a new Sequential optimization framework for healthcare process optimizations in this work. It is a machine learning tool that provides equivalent human actions by deciding to choose certain future moves independently. Results for the second experiment: SFM visually explains process flow and capiases dependency between different steps in a sort order. These two techniques are combined in the proposed framework to learn and optimize healthcare processes sequentially. Sequential optimization: the RL algorithms started learning from previous experiences to optimize the best decisions at different process steps. From there, the SFM provides a visual rendering of this streamlined process. The effect of the process changes can be experimented with via simulation and altered as such. This iterative process improves efficiency in processes and reduces costs to healthcare systems. The proposed framework may transform healthcare processes' optimization and, by extension, improve patient outcomes.