Modelling Arterial Pressure Waveforms Using Gaussian Functions and Two-Stage Particle Swarm Optimizer

Table 2

Statistical results of MAE from 20 volunteers with different max_FEs values for the four methods.

max_FEs	CAPW (%)				RAPW (%)
	Nelder-Mead	MPSO	DMS-PSO	TSPSO	Nelder-Mead	MPSO	DMS-PSO	TSPSO
2000	9.8±1.9^∗	7.0±1.2	7.5±1.3^∗	6.4±1.0	8.1±1.7^∗	5.7±1.0	6.3±1.1^∗	5.2±0.9
5000	6.2±1.4^∗	5.6±0.9^∗	5.3±1.0^∗	4.6±0.7	6.0±1.4^∗	4.5±0.9^∗	4.0±0.7	3.9±0.6
10000	4.7±1.1^∗	4.5±0.7^∗	2.6±0.6^∗	2.1±0.5	4.5±1.2^∗	4.0±0.6^∗	2.5±0.5^∗	1.5±0.4
15000	4.4±1.0^∗	3.4±0.6^∗	2.2±0.5^∗	1.7±0.4	4.3±1.0^∗	2.8±0.6^∗	1.5±0.4	1.3±0.4
20000	4.3±0.9^∗	2.6±0.5^∗	1.5±0.3	1.4±0.3	4.2±0.8^∗	2.4±0.5^∗	1.4±0.3^∗	1.2±0.2
25000	4.2±0.8^∗	2.1±0.3^∗	1.3±0.2	1.2±0.2	4.1±0.7^∗	2.0±0.3^∗	1.2±0.2	1.1±0.2
30000	4.2±0.7^∗	2.0±0.3^∗	1.2±0.2	1.1±0.2	4.1±0.6^∗	1.9±0.3^∗	1.1±0.2	1.0±0.2

Data are expressed as mean ± standard deviation (SD). MAE: mean absolute error; max_FEs: maximum number of function evaluations; CAPW: carotid artery pressure waveforms; RAPW: radial artery pressure waveforms.
∗There is statistically significant difference when compared with TSPSO (𝑃<0.05).