A Deep Learning Approach to Optimal Sampling Problems

Table 1

Parameters’ value.

System and control	Parameters	Values
Neural networks	Hidden layers	1
	Number of neurons	5
Soft greedy policy	Exploration degree	0.3
2D system, periodic impulse control	Pole of the system	0
	Weight parameter	2
	Steps	1000
	Batch	10
	Integer	200
	Precision
	The range of
2D system, event-based impulse control	Pole of the system	0
	Weight parameter	2
	Steps	1000
	Batch	10
	Integer	200
	Precision
	The range of
3D system, periodic impulse control	Matrix
	Weight parameter	2
	Steps	1000
	Batch	10
	Integer	200
	Precision
	The range of
3D system, event-based impulse control	Matrix
	Weight parameter	2
	Steps	1000
	Batch	10
	Integer	200
	Precision
	The range of