Degree-Constrained -Minimum Spanning Tree Problem

Table 13

Numerical results obtained with Q-learning algorithm for random and Euclidean instances using .

#
			Best	Time (s)			Time (s)	#Iter
Input graphs with random uniform costs
1	50	10	52.22	0.61	427.26	52.22	9.03	42359	0
2		20	134.08	1.61	794.00	138.00	8.76	40622	2.92
3		30	240.30	1.08	1473.09	245.62	2.25	10079	2.21
4		50	564.37	0.47	2567.98	601.37	15.38	70922	6.55
5	100	25	123.71	2.95	1207.44	123.71	4.75	9890	0
6		50	287.78	6.04	2440.56	291.89	98.12	211891	1.42
7		75	529.56	0.84	3473.78	565.78	487.90	1067899	6.83
8		100	915.29	3.34	4975.56	1024.77	30.10	64345	11.96
9	200	50	184.29	4.04	2543.82	199.27	130.44	122671	8.12
10		100	437.08	30.84	5027.12	476.04	123.87	118105	8.91
11		150	775.89	18.53	7342.28	843.60	38.57	36512	8.72
12		200	1285.32	6.43	9799.77	1546.62	227.21	214107	20.32
Input graphs with Euclidean distance costs
1	Berlin52	10	279.23	0.51	2692.71	279.23	39.93	200322	0
2		20	1015.39	2.09	12651.13	1020.02	151.28	774386	0.45
3		30	2065.12	2.23	16299.43	2070.65	45.42	228363	0.26
4		52	6968.77	0.73	30795.84	7234.27	85.36	433901	3.80
5	gr96	25	73.56	621.64	858.80	74.59	78.73	199419	1.40
6		50	168.53	299.41	1774.49	180.05	89.15	225194	6.83
7		75	283.14	96.28	2500.70	303.35	260.64	648302	7.13
8		96	474.95	17.58	3352.78	552.06	191.68	477430	16.23
9	ch150	50	1753.83	3600	18911.28	1711.33	125.94	186763	−2.42
10		80	2881.72	3600	28695.92	2949.71	202.88	306120	2.35
11		100	3627.93	279.88	37580.42	3953.79	107.61	160624	8.98
12		150	6368.79	43.71	54468.42	7122.77	565.44	844020	11.83