Table 6 Comparison of prediction model results with/without parallel mechanism under three timeframe partition methods.

Data	Timeframe	Disjoint		Overlapping		SATPM
	Timeframe	size:1	offset:1	size:2	offset:1	/
	parallel mechanism	without	with	without	with	without	with
AS	\(\overline{Acc}\)	0.916	0.928	0.923	0.922	0.913	0.913
	Macro avg	0.259	0.605	0.316	0.477	0.423	0.439
	Weighted avg	0.889	0.920	0.908	0.896	0.893	0.894
AS-Caida	\(\overline{Acc}\)	0.926	0.903	0.900	0.924	0.965	0.978
	Macro avg	0.885	0.834	0.879	0.884	0.891	0.937
	Weighted avg	0.925	0.902	0.898	0.920	0.967	0.980
DBLP	\(\overline{Acc}\)	0.908	0.790	0.884	0.896	0.986	0.994
	Macro avg	0.801	0.593	0.870	0.887	0.983	0.996
	Weighted avg	0.902	0.767	0.885	0.895	0.986	0.994
Facebook	\(\overline{Acc}\)	0.802	0.681	0.934	0.946	0.768	0.850
	Macro avg	0.756	0.694	0.904	0.909	0.590	0.767
	Weighted avg	0.802	0.707	0.930	0.951	0.737	0.829
Sx-askubuntu-c2q	\(\overline{Acc}\)	0.945	0.906	0.969	0.981	0.975	0.987
	Macro avg	0.930	0.841	0.907	0.964	0.922	0.951
	Weighted avg	0.945	0.904	0.962	0.980	0.972	0.989

Quick links

Search