Ensemble Just-In-Time Learning-Based Soft Sensor for Mooney Viscosity Prediction in an Industrial Rubber Mixing Process

Table 1

Characteristics of soft sensor methods for comparison.

No.	Method	Model structure	Learning type	Diversity generation mechanism
1	PLS	Single	Global	—
2	GPR	Single	Global	—
3	GMMGPR	Ensemble	Local	Training data perturbation
4	JITGPR (ED similarity)	Single	Local	—
5	JITGPR (cosine similarity)	Single	Local	—
6	JITGPR (CWD similarity)	Single	Local	—
7	JITGPR (CC similarity)	Single	Local	—
8	SP-EJITGPR (SAR)	Ensemble	Local	Similarity perturbation
9	SP-EJITGPR (PLS stacking)	Ensemble	Local	Similarity perturbation
10	SP-EJITGPR (GPR stacking)	Ensemble	Local	Similarity perturbation
11	SP-EJITGPR (FMM)	Ensemble	Local	Similarity perturbation
12	MP-EJITGPR (SAR)	Ensemble	Local	Multimodal perturbation
13	MP-EJITGPR (PLS stacking)	Ensemble	Local	Multimodal perturbation
14	MP-EJITGPR (GPR stacking)	Ensemble	Local	Multimodal perturbation
15	MP-EJITGPR (FMM)	Ensemble	Local	Multimodal perturbation