reinforcement learning has been increasingly applied in monitoring
applications because of its ability to learn from previous experiences and can
make adaptive decisions. However, existing machine learning-based health
monitoring applications are mostly supervised learning algorithms,