Network Security Defense Decision-Making Method Based on Stochastic Game and Deep Reinforcement Learning

Table 3

Comparison between the method in this article and the typical methods.

References	Theory	Game type	Game information	Behavioral rationality	Transition probability	Equilibrium solution	Accuracy	Versatility
[8]	Stochastic game	Static	Complete information	Completely rational	Fixed value	Simple	Poor	Poor
[9]	Stochastic evolutionary game	Dynamic	Incomplete information	Bounded rationality	Dynamic changes	Detailed	Good	Good
[28]	Stochastic game + reinforcement learning	Dynamic	Incomplete information	Bounded rationality	Dynamic changes	Detailed	Good	Good
Our method	Stochastic game + deep reinforcement learning	Dynamic	Incomplete information	Bounded rationality	Dynamic changes	Detailed	Better	Better