A multi-country comparison of stochastic models of breast cancer mortality with P-splines smoothing approach

Table 1 Error measures from three fitted models individually for each country

	Smoothed Lee Carter (SLC) Model
Average across ages	ME	MSE	MPE	MAPE
China	0.000	0.0008	0.020	0.036
Pakistan	0.000	0.0024	0.0004	0.016
India	0.000	0.0007	0.0232	0.073
Thailand	0.000	0.0023	0.0110	0.056
Average across years
China	0.000	0.046	1.509	2.736
Pakistan	0.000	0.134	0.018	0.794
India	0.000	0.0418	1.7075	5.827
Thailand	0.000	0.133	0.803	4.153
Average across ages	Functional Demographic Model (FDM)
China	0.000	0.0013	−0.029	0.065
Pakistan	−0.000	0.0004	0.001	0.008
India	0.000	0.0003	0.0128	0.046
Thailand	−0.000	0.000	0.005	0.019
Average across years
China	−0.000	0.0700	−2.199	4.815
Pakistan	−0.000	0.026	0.0437	0.367
India	−0.000	0.016	0.939	3.575
Thailand	−0.000	0.015	0.378	1.359
Average across ages	Booth-Maindonald-Smith (BMS) Model
China	−0.000	0.0014	0.033	0.054
Pakistan	0.001	0.002	0.001	0.016
India	0.000	0.001	0.022	0.075
Thailand	−0.000	0.006	0.032	0.141
Average across years
China	−0.007	0.084	2.466	4.109
Pakistan	0.077	0.139	0.048	0.788
India	−0.000	0.045	1.675	5.909
Thailand	−0.004	0.362	2.348	10.701

ME mean error, MSE mean square error; MPE mean percent error; MAPE mean absolute percent error

ISSN: 1471-2288