Research on Calibration Transfer Method Based on Joint Feature Subspace Distribution Alignment

Yu-hui ZHAO; Xiao-dong LIU; Lei ZHANG; Yong-hong LIU

doi:10.3964/j.issn.1000-0593(2021)11-3411-07

Journals >Spectroscopy and Spectral Analysis >Volume 41 >Issue 11 >Page 3411 > Article

Spectroscopy and Spectral Analysis
Vol. 41, Issue 11, 3411 (2021)

Research on Calibration Transfer Method Based on Joint Feature Subspace Distribution Alignment

Yu-hui ZHAO^*, Xiao-dong LIU, Lei ZHANG, and Yong-hong LIU

Author Affiliations

Northeastern University Qinhuangdao Campus, Qinhuangdao 066000, China

show less

DOI: 10.3964/j.issn.1000-0593(2021)11-3411-07 Cite this Article

Yu-hui ZHAO, Xiao-dong LIU, Lei ZHANG, Yong-hong LIU. Research on Calibration Transfer Method Based on Joint Feature Subspace Distribution Alignment[J]. Spectroscopy and Spectral Analysis, 2021, 41(11): 3411 Copy Citation Text

show less

Fig. 1. Feature distribution alignment diagram
(a): Original; (b): Mean correction; (c): Covariance correction

Download full size | View in the Article

Fig. 2. Spectral differences between different instruments

Download full size | View in the Article

Fig. 3. Scatter plots of prediction comparison between instruments M5 and MP5 using JSDA, SBC, PDS, CCACT, MSC, TCR

Download full size | View in the Article

Fig. 4. Scatter plots of prediction comparsion between instruments M5 and MP6 using JSDA, SBC, PDS, CCACT, MSC, TCR

Download full size | View in the Article

Fig. 5. Scatter plots of prediction comparison between instruments MP5 and MP6 using JSDA, SBC, PDS, CCACT, MSC, TCR

Download full size | View in the Article

Fig. 6. Scatter plots of prediction comparison between instruments A1 and A2 using JSDA, SBC, PDS, CCACT, MSC, TCR

Download full size | View in the Article

Fig. 7. Scatter plots of prediction comparison between instruments A1 and A3 using JSDA, SBC, PDS, CCACT, MSC, TCR

Download full size | View in the Article

Fig. 8. Scatter plots of prediction comparison between instruments A3 and A2 using JSDA, SBC, PDS, CCACT, MSC, TCR

Download full size | View in the Article

N_std	SBC	PDS	CCACT	MSC	TCR	JSDA
玉米数据集的RMSEP (M5作为主仪器, MP5作为从仪器)
N=15	0.317 49	0.240 54(15^a)	0.252 59
N=25	0.251 58	0.218 47(15^a)	0.250 05	0.899 59	0.355 51(6^b)	0.118 64
N=35	0.267 10	0.227 79(15^a)	0.246 27
玉米数据集的RMSEP (M5作为主仪器, MP6作为从仪器)
N=15	0.436 55	0.433 16(7^a)	0.546 91
N=25	0.388 60	0.467 49(9^a)	0.420 87	1.921 09	0.474 99(4^b)	0.146 09
N=35	0.355 98	0.403 44(5^a)	0.406 74
玉米数据集的 RMSEP (MP5作为主仪器, MP6作为从仪器)
N=15	0.225 77	0.240 54(15^a)	0.252 59
N=25	0.227 63	0.218 47(15^a)	0.250 05	0.899 59	0.355 51(10^b)	0.172 82
N=35	0.238 45	0.227 79(15^a)	0.246 27

Table 1. RMSEP of corn datasets with SBC, PDS, CCACT, MSC, TCR and JSDA

View in the Article

N_std	SBC	PDS	CCACT	MSC	TCR	JSDA
小麦数据集的RMSEP (A1作为主仪器, A2作为从仪器)
N=15	2.170 63	3.947 74(15^a)	2.322 63
N=25	1.284 26	3.793 84(15^a)	2.046 23	1.605 83	2.035 40(10^b)	0.283 44
N=35	0.905 77	3.504 91(15^a)	1.988 81
小麦数据集的RMSEP (A1作为主仪器, A3作为从仪器)
N=15	0.477 07	1.524 29(7^a)	2.030 12
N=25	0.477 39	1.408 40(15^a)	2.071 13	1.215 13	1.964 12(19^b)	0.252 32
N=35	0.478 35	1.346 55(15^a)	1.902 66
小麦数据集的RMSEP (A3作为主仪器, A2作为从仪器)
N=15	8.660 48	3.593 11(15^a)	2.249 79
N=25	7.082 65	2.229 61(3^a)	2.250 17	1.255 70	2.120 74(10^b)	0.277 81
N=35	5.996 79	2.069 79(3^a)	2.062 74

Table 2. RMSEP of wheat datasets with SBC, PDS, CCACT, MSC, TCR and JSDA

Download Citation

Save the article for my favorites

Paper Information