A Novel Self-Adaptive Harmony Search Algorithm

Table 3

The descriptive statistical result of the relative errors of the NSHS algorithm in the case of HMS = 5, 10, and 50, respectively.

Function	HMS = 5		HMS = 10		HMS = 20		HMS = 50
	Mean	Std.	Mean	Std.	Mean	Std.	Mean	Std.
f1	7.54𝐸−07	4.24𝐸−07	4.65𝐸−07	2.53𝐸−07	1.00𝐸−06	5.24𝐸−07	5.37𝐸−06	1.69𝐸−06
f2	2.96𝐸−01	5.36𝐸−01	1.56𝐸−01	1.45𝐸−01	1.22𝐸−01	1.36𝐸−01	1.09𝐸−01	1.32𝐸−01
f3	2.74𝐸−06	1.72𝐸−06	4.92𝐸−06	3.28𝐸−06	1.48𝐸−03	8.05𝐸−03	1.98𝐸−03	8.55𝐸−03
f4	3.76𝐸−05	1.76𝐸−05	4.71𝐸−05	1.32𝐸−05	7.23𝐸−05	1.64𝐸−05	1.34𝐸−04	2.53𝐸−05
f5	1.32𝐸−04	5.49𝐸−05	1.41𝐸−04	3.03𝐸−05	2.09𝐸−04	4.82𝐸−05	4.17𝐸−04	8.23𝐸−05
f6	4.22𝐸−07	2.46𝐸−07	1.11𝐸−07	6.73𝐸−08	3.27𝐸−07	2.02𝐸−07	1.81𝐸−06	6.08𝐸−07
f7	3.38𝐸−06	1.46𝐸−06	6.21𝐸−06	2.01𝐸−06	1.07𝐸−05	3.03𝐸−06	2.07𝐸−05	4.22𝐸−06
f8	2.25𝐸−09	9.12𝐸−10	2.95𝐸−09	1.26𝐸−09	2.90𝐸−08	1.80𝐸−08	7.47𝐸−07	2.82𝐸−07
f9	8.00𝐸−05	8.44𝐸−05	1.07𝐸−04	7.34𝐸−05	9.62𝐸−05	6.45𝐸−05	3.74𝐸−05	3.95𝐸−05
f10	2.12𝐸−03	2.69𝐸−03	1.10𝐸−03	1.69𝐸−03	5.03𝐸−04	1.39𝐸−03	3.59𝐸−04	9.48𝐸−04
f11	3.73𝐸−04	4.75𝐸−05	3.81𝐸−04	4.29𝐸−05	4.76𝐸−04	3.89𝐸−05	6.59𝐸−04	5.76𝐸−05
f12	3.59𝐸−07	2.67𝐸−07	1.26𝐸−07	9.95𝐸−08	2.16𝐸−07	1.71𝐸−07	9.81𝐸−07	4.43𝐸−07
f13	2.08𝐸−04	1.61𝐸−04	2.77𝐸−04	2.07𝐸−04	3.24𝐸−04	1.79𝐸−04	5.51𝐸−04	3.38𝐸−04
f14	3.52𝐸−06	4.29𝐸−06	3.36𝐸−06	2.67𝐸−06	7.43𝐸−06	4.40𝐸−06	9.78𝐸−06	5.48𝐸−06
f15	2.47𝐸−05	6.67𝐸−05	4.60𝐸−05	1.91𝐸−04	7.40𝐸−05	1.95𝐸−04	1.86𝐸−03	1.56𝐸−03
f16	1.17𝐸−04	7.96𝐸−05	1.26𝐸−04	8.46𝐸−05	1.16𝐸−04	6.32𝐸−05	1.75𝐸−04	7.27𝐸−05
f17	2.89𝐸−02	2.92𝐸−02	2.09𝐸−02	2.72𝐸−02	2.63𝐸−02	2.48𝐸−02	2.96𝐸−02	3.21𝐸−02
f18	1.44𝐸−02	9.70𝐸−03	7.29𝐸−03	3.81𝐸−03	9.57𝐸−03	3.65𝐸−03	1.50𝐸−02	6.03𝐸−03
Total	1.90𝐸−02	1.42𝐸−01	1.04𝐸−02	4.95𝐸−02	8.96𝐸−03	4.28𝐸−02	8.85𝐸−03	4.05𝐸−02

∗Each group has 30 samples and the experiment has totally 4 * 18 * 30 = 2160 samples.