Limits...
A non-destructive distinctive method for discrimination of automobile lubricant variety by visible and short-wave infrared spectroscopy.

Jiang L, Liu F, He Y - Sensors (Basel) (2012)

Bottom Line: In order to improve prediction precision further, BVB was processed by UVE-PLS and the optimal cutoff threshold of UVE was found by SA.An optimal model with a correlation coefficient (R) of 0.9850 and root mean square error of prediction (RMSEP) of 0.0827 was obtained.The overall results indicated that the method of combining WPT, UVE-PLS and SA was a powerful way to select diagnostic information for discrimination among different varieties of lubricating oil, furthermore, a more parsimonious and efficient LS-SVM model could be obtained.

View Article: PubMed Central - PubMed

Affiliation: Zhejiang Technology Institute of Economy, Hangzhou 310018, China. abbyu111@sina.com

ABSTRACT
A novel method which is a combination of wavelet packet transform (WPT), uninformative variable elimination by partial least squares (UVE-PLS) and simulated annealing (SA) to extract best variance information among different varieties of lubricants is presented. A total of 180 samples (60 for each variety) were characterized on the basis of visible and short-wave infrared spectroscopy (VIS-SWNIR), and 90 samples (30 for each variety) were randomly selected for the calibration set, whereas, the remaining 90 samples (30 for each variety) were used for the validation set. The spectral data was split into different frequency bands by WPT, and different frequency bands were obtained. SA was employed to look for the best variance band (BVB) among different varieties of lubricants. In order to improve prediction precision further, BVB was processed by UVE-PLS and the optimal cutoff threshold of UVE was found by SA. Finally, five variables were mined, and were set as inputs for a least square-support vector machine (LS-SVM) to build the recognition model. An optimal model with a correlation coefficient (R) of 0.9850 and root mean square error of prediction (RMSEP) of 0.0827 was obtained. The overall results indicated that the method of combining WPT, UVE-PLS and SA was a powerful way to select diagnostic information for discrimination among different varieties of lubricating oil, furthermore, a more parsimonious and efficient LS-SVM model could be obtained.

No MeSH data available.


The absolute deviation value of prediction results of BVB-RV-LS-SVM.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3376611&req=5

f8-sensors-12-03498: The absolute deviation value of prediction results of BVB-RV-LS-SVM.

Mentions: The retained five variables constituted a new set of variables for LS-SVM (BVB-RV-LS-SVM). The performance of BVB-RV-LS-SVM was evaluated by the same 90 unknown samples mentioned above, and R of 0.9950 and RMSEP of 0.0829 were obtained, respectively. The absolute deviation of prediction results of each sample is shown in Figure 8. As one could see from Figure 8, when the threshold error was set as ±0.3, no sample was mistakenly discriminated, if the threshold error was ±0.2, there were only two misclassified samples, these were samples 62 and 65, and the total discrimination rate reached 97.78%. Therefore, compared with BVB-LS-SVM, the BVB-RV-LS-SVM model not only improved the prediction accuracy greatly, but also produced a more parsimonious model.


A non-destructive distinctive method for discrimination of automobile lubricant variety by visible and short-wave infrared spectroscopy.

Jiang L, Liu F, He Y - Sensors (Basel) (2012)

The absolute deviation value of prediction results of BVB-RV-LS-SVM.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3376611&req=5

f8-sensors-12-03498: The absolute deviation value of prediction results of BVB-RV-LS-SVM.
Mentions: The retained five variables constituted a new set of variables for LS-SVM (BVB-RV-LS-SVM). The performance of BVB-RV-LS-SVM was evaluated by the same 90 unknown samples mentioned above, and R of 0.9950 and RMSEP of 0.0829 were obtained, respectively. The absolute deviation of prediction results of each sample is shown in Figure 8. As one could see from Figure 8, when the threshold error was set as ±0.3, no sample was mistakenly discriminated, if the threshold error was ±0.2, there were only two misclassified samples, these were samples 62 and 65, and the total discrimination rate reached 97.78%. Therefore, compared with BVB-LS-SVM, the BVB-RV-LS-SVM model not only improved the prediction accuracy greatly, but also produced a more parsimonious model.

Bottom Line: In order to improve prediction precision further, BVB was processed by UVE-PLS and the optimal cutoff threshold of UVE was found by SA.An optimal model with a correlation coefficient (R) of 0.9850 and root mean square error of prediction (RMSEP) of 0.0827 was obtained.The overall results indicated that the method of combining WPT, UVE-PLS and SA was a powerful way to select diagnostic information for discrimination among different varieties of lubricating oil, furthermore, a more parsimonious and efficient LS-SVM model could be obtained.

View Article: PubMed Central - PubMed

Affiliation: Zhejiang Technology Institute of Economy, Hangzhou 310018, China. abbyu111@sina.com

ABSTRACT
A novel method which is a combination of wavelet packet transform (WPT), uninformative variable elimination by partial least squares (UVE-PLS) and simulated annealing (SA) to extract best variance information among different varieties of lubricants is presented. A total of 180 samples (60 for each variety) were characterized on the basis of visible and short-wave infrared spectroscopy (VIS-SWNIR), and 90 samples (30 for each variety) were randomly selected for the calibration set, whereas, the remaining 90 samples (30 for each variety) were used for the validation set. The spectral data was split into different frequency bands by WPT, and different frequency bands were obtained. SA was employed to look for the best variance band (BVB) among different varieties of lubricants. In order to improve prediction precision further, BVB was processed by UVE-PLS and the optimal cutoff threshold of UVE was found by SA. Finally, five variables were mined, and were set as inputs for a least square-support vector machine (LS-SVM) to build the recognition model. An optimal model with a correlation coefficient (R) of 0.9850 and root mean square error of prediction (RMSEP) of 0.0827 was obtained. The overall results indicated that the method of combining WPT, UVE-PLS and SA was a powerful way to select diagnostic information for discrimination among different varieties of lubricating oil, furthermore, a more parsimonious and efficient LS-SVM model could be obtained.

No MeSH data available.