Identification of a Typical CSTR Using Optimal Focused Time Lagged Recurrent Neural Network Model with Gamma Memory Filter

Table 5

Comparison of NN models.

NN Models	Performance Measures on Test Data Set
	CSTR lab data			Benchmark CSTR data
	MSE	NMSE	𝑟	MSE	NMSE	𝑟
FTLR NN with Gamma memory	1.13E−05	0.470	0.9132	1.40E−06	0.00777	0.9961
Fully Recurrent NN	8.01E−06	0.754	0.702	4.15E−06	0.0230	0.988