Table 1.

Examples of NP-hard graph problems that are defined in RL algorithms

Problem	State	Action	Reward	Termination
MVC	Subset of nodes selected as the partial solution	Add a new node to the partial solution	Number of nodes used to cover all edges at the end of the episode	All edges are covered
MAX	Subset of nodes selected as the partial solution	Add a new node to the partial solution	Cut set weight at the end of the episode	Cut set weight cannot be improved