Event Forecasting in Organizational Networks: A Discrete Dynamical System Approach

Table 3

Aggregated statistics of the 237 iterations which produced the highest reliability accuracy metrics. Source: own elaboration.

		Window size
		24 weeks	16 weeks	8 weeks	Mean
Prediction accuracy (p)	Mean	0.864547736	0.88407903	0.877279795	0.873421925
	Standard deviation	0.089712353	0.063456924	0.096425389	0.08341138
Test datasets	Mean	905.9298246	1353.316456	2639.863636	1376.970464
	Standard deviation	93.64645615	177.4148799	423.4860636	672.3086483
Training datasets	Mean	2921.535088	3882.329114	7452.909091	4083.067511
	Standard deviation	1595.176985	1149.72304	1689.871563	2226.650539
Test dataset’s output class ratio (R)	Mean	0.493061243	0.485309883	0.48420594	0.488833434
	Standard deviation	56.20898166	38.33948073	21.30506137	115.8535238
Training dataset’s output class ratio (R)	Mean	0.325785796	0.277650833	0.267522719	0.298924035
	Standard deviation	0.237382294	0.19001084	0.174750185	0.212513446