Research on Amplifier Performance Evaluation Based on δ-Support Vector Regression

Table 1

Result of comparative experiment.

TRSN	TESN	FN	Method	Parameter; (𝜎,𝜉,𝐶)	SVN	TRMSE	TEMSE
259×100	59×100	8	𝛿-SVR	(0.1, 0.01, 6)	1095	2.4291𝑒−026	2.0009𝑒−031
259×100	59×100	8	LSSVR	(0.1, 0.01, 6)	1689	3.3619𝑒−019	3.9276𝑒−023
259×100	59×100	8	𝜀-SVR	(0.1, 0.01, 6)	1407	2.4967𝑒−018	2.0138𝑒−015

Note  2. SVN denotes the number of support vectors, TESN denotes the number of testing support vectors, TRSN denotes the number of training support vectors, FN denotes the number of the data features, TEMSE denotes the testing data mean square error and TDMSE denotes the training data mean square error.