A New Separable Piecewise Linear Learning Algorithm for the Stochastic Empty Container Repositioning Problem

Table 8

Percent deviation between the PB and the other algorithms on the operational costs.

Problem	Algorithm	Number of iterations					Computation time/iteration (s)
		40	160	640	1200	4000
P1	SPELL	9.28	3.56	0.71	0.28	0.07	0.02
	L-shaped	1.17	0.85	0.05	0.05	0.05	0.09
	Myopic	3.45	2.78	2.78	2.78	2.78	—
P2	SPELL	6.12	1.02	0.34	0.07	0.05	0.31
	L-shaped	4.57	2.23	0.31	0.06	0.04	3.58
	Myopic	7.68	7.21	7.15	7.15	7.15	—
P3	SPELL	4.68	0.76	0.13	0.05	0.02	1.93
	L-shaped	5.97	3.28	1.38	0.62	0.36	22.90
	Myopic	8.21	7.66	7.63	7.61	7.61	—