Using Consensus Strategy and Interval Partial Least Square Algorithm in Wavelet Domain for Analysis of Near-infrared Spectroscopy
Dan Peng, Guo He, Linqing Li, Yanlan Bi, Guolong Yang
Available Online June 2018.
- https://doi.org/10.2991/icmmct-18.2018.22How to use a DOI?
- consensus strategy; interval partial least square algorithm; near-infrared spectroscopy; wavelet domain; selection; integration
- To improve the stability and precision performance of partial least square regression (PLS) model in near-infrared analysis application, the consensus strategy was applied in the wavelet domain. Taking the advantage of multiscale property of wavelet packet analysis, a new modelling method was developed based on the idea of the interval PLS algorithm and named as WpCo-iPLS algorithm. In WpCo-iPLS model, wavelet packet transform (WPT) algorithm was firstly adopted to split the raw spectra into a series of frequency components in wavelet domain. Then, coupled with the consensus strategy, multiple member PLS models were established on the interval frequency components. To reduce the dependence on single model, an optimization of the weight parameters of member models was conducted. At last, a consensus model was achieved by effectively combining all the member models. To validate the WpCo-iPLS algorithm, it was applied to measure the six kinds of contents concentration of diesel samples using NIR spectra. The experimental results showed that the prediction ability and robustness of WpCo-iPLS model was stronger than that of conventional consensus algorithms, indicating that it is a promising consensus strategy for modelling using NIR spectra.
- Open Access
- This is an open access article distributed under the CC BY-NC license.
Cite this article
TY - CONF AU - Dan Peng AU - Guo He AU - Linqing Li AU - Yanlan Bi AU - Guolong Yang PY - 2018/06 DA - 2018/06 TI - Using Consensus Strategy and Interval Partial Least Square Algorithm in Wavelet Domain for Analysis of Near-infrared Spectroscopy BT - 2018 6th International Conference on Machinery, Materials and Computing Technology (ICMMCT 2018) PB - Atlantis Press SP - 113 EP - 119 SN - 2352-5401 UR - https://doi.org/10.2991/icmmct-18.2018.22 DO - https://doi.org/10.2991/icmmct-18.2018.22 ID - Peng2018/06 ER -