Exploring Cooperative Game Mechanisms of Scientific Coauthorship Networks

Table 5

The parameters and goodness of fitting functions for degree distribution, hyperdegree distribution, and the distribution of paper team sizes, from top to down.

								value
PNAS	4.751	0.434	110.8	2.987	0.845	10	25	0.256
	0.085	0.367	2.368	2.941	11.99	3	7	0.128
	3.438	0.382	3,094	4.755	1.095	16	17	0.174
Synthetic-1	5.082	0.380	708.4	3.580	0.898	13	32	0.074
	0.664	0.365	16.15	3.694	1.999	4	9	0.971
	4.502	0.208	3,621	5.059	1.003	6	20	0.103
PRE	2.776	0.156	8.910	2.758	0.898	5	6	0.118
	0.038	0.399	4.260	3.079	25.56	3	10	0.875
	2.507	0.000	82.66	4.587	1.249	6	7	0.133
Synthetic-2	2.352	0.2367	46.65	3.460	0.970	5	11	0.052
	0.375	0.272	1.786	2.991	3.077	5	6	0.673
	2.007	0.065	116.1	4.756	1.146	5	7	0.887

The domains of generalized Poisson , crossover, and power law are , and , respectively. The fitting function defined on is where . The fitting processes are the following: calculate parameters of and by regressing the head and tail of empirical distribution, respectively and find and through exhaustion to make pass the KS test ( value > 0.05). The value measures the goodness-of-fit.