Using Sequence Mining to Predict Complex Systems: A Case Study in Influenza Epidemics

Table 4

Comparison of different models with the INTS model for the prediction of influenza epidemics using Google search queries.

Models	MSE	RMSE	MAE	Correlation of increment
INTS model (our model)	0.0013	0.0369	0.0185	0.9718
ARGO model [49]	0.3697	0.608	0.649	0.758
GFT	4.911	2.216	1.834	0.706
GFT + AR (3) [12]	0.8372	0.915	1.052	0.690
AR (3) model [9]	0.9157	0.957	0.925	0.385
Naive model [9]	0.1211	0.348	0.201	0.436
ANNPSO (proposed model)	0.0024	0.049	0.25	0.94
SVMR (proposed model)	0.136	0.369	3.88	0.99