Weighted Fusion of Multiple Models for Wavelength Selection

Zeng, Tian-Ling; Zhang, Wen; Wen, Zhongquan; Hong, Mingjian

doi:10.1366/12-06757

Cited by 4 publications

(3 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We thought that the spectral regions extracted with the CA were subjective, and some redundant and inaccurate spectral information might be occurred in the predictive model, which will definitely produce negative effect on the model performance. For SPA+MLR, the spectral wavelengths were directly and initially determined and then were used to establish the monitor model, and this may lead to the fact that more redundant and irrelevant spectral information will negatively contribute the model performance (Soares et al., 2013; Zeng, Wen, Wen, & Hong, 2013). However, for the PLSR+SMLR, the important spectral regions which were highly related with the ChD of winter wheat were determined by using the combination of the VIP and B‐coefficient derived from the PLSR analysis, then the SMLR were further applied to selected the sensitive wavelengths which proved to be important to the ChD.…”

Section: Discussionmentioning

confidence: 99%

Hyperspectral monitor on chlorophyll density in winter wheat under water stress

et al. 2020

View full text Add to dashboard Cite

Lack of water can lead lower chlorophyll concentrations and yields. Based on the hyperspectral reflectance of winter wheat (Triticum aestivum L.), we analyzed the relationship between the canopy hyperspectral reflectance and the waterstressed winter wheat. The correlation analysis (CA), partial least squares regression (PLSR), and successive progressions algorithm (SPA) were used to extract the important bands. Then the hyperspectral estimation models for chlorophyll density (ChD) by characteristic variables were established. The results showed that the reflectance in the visible regions increased gradually with an increasing water stress. In the near-infrared (NIR) region, reflectance decreased with stress intensity. We extracted five and seven important waveband regions by CA and PLSR. Then we extracted right and nine important bands through SMLR, and 11 important bands were extracted by the method of SPA. We found that 427, 434, 749, and 814 nm contained important information about ChD of winter wheat after water stress. The model established by CA+SMLR was generally realized, whereas the ChD estimation models established by PLSR+SMLR and SPA with multiple linear regression had better performance, and the performance of the validation model was accurate and robust. The results of this study could provide theoretical basis and practical reference for accurate estimation of ChD in winter wheat after water stress.

show abstract

Section: Discussionmentioning

confidence: 99%

Hyperspectral monitor on chlorophyll density in winter wheat under water stress

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Based on the advantages of reversible-jump Markov Chain Monte Carlo, a random frog wavelength selection approach is proposed and implemented easily [18]. A new wavelength selection method based on the weighted fusion of multiple models is proposed [20], where the different PLS submodels are built with different numbers of latent variables, and then the regression coefficients of multiple PLS models are fused to select the wavelengths. A new wavelength selection method based on the weighted fusion of multiple models is proposed [20], where the different PLS submodels are built with different numbers of latent variables, and then the regression coefficients of multiple PLS models are fused to select the wavelengths.…”

Section: Introductionmentioning

confidence: 99%

“…A wavelength selection algorithm based on exponentially weighted recursive PLS and variance importance in the projection is discussed [19]. A new wavelength selection method based on the weighted fusion of multiple models is proposed [20], where the different PLS submodels are built with different numbers of latent variables, and then the regression coefficients of multiple PLS models are fused to select the wavelengths. A rough set-based wavelength selection method is presented in NIR spectral analysis [21].…”

Section: Introductionmentioning

confidence: 99%

A wavelength selection method based on random decision particle swarm optimization with attractor for near‐infrared spectral quantitative analysis

Cao

Wang

Yang

et al. 2015

Journal of Chemometrics

View full text Add to dashboard Cite

In this paper, we proposed a wavelength selection method based on random decision particle swarm optimization with attractor for near-infrared (NIR) spectra quantitative analysis. The proposed method was incorporated with partial least square (PLS) to construct a prediction model. The proposed method chooses the current own optimal or the current global optimal to calculate the attractor. Then the particle updates its flight velocity by the attractor, and the particle state is updated by the random decision with the new velocity. Moreover, the root-mean-square error of cross-validation is adopted as the fitness function for the proposed method. In order to demonstrate the usefulness of the proposed method, PLS with all wavelengths, uninformative variable elimination by PLS, elastic net, genetic algorithm combined with PLS, the discrete particle swarm optimization combined with PLS, the modified particle swarm optimization combined with PLS, the neighboring particle swarm optimization combined with PLS, and the proposed method are used for building the components quantitative analysis models of NIR spectral datasets, and the effectiveness of these models is compared. Two application studies are presented, which involve NIR data obtained from an experiment of meat content determination using NIR and a combustion procedure. Results verify that the proposed method has higher predictive ability for NIR spectral data and the number of selected wavelengths is less. The proposed method has faster convergence speed and could overcome the premature convergence problem. Furthermore, although improving the prediction precision may sacrifice the model complexity under a certain extent, the proposed method is overfitted slightly.

show abstract

A consensus successive projections algorithm – multiple linear regression method for analyzing near infrared spectra

Liu

Chen

et al. 2015

Analytica Chimica Acta

View full text Add to dashboard Cite

Weighted Fusion of Multiple Models for Wavelength Selection

Cited by 4 publications

References 25 publications

Hyperspectral monitor on chlorophyll density in winter wheat under water stress

Hyperspectral monitor on chlorophyll density in winter wheat under water stress

A wavelength selection method based on random decision particle swarm optimization with attractor for near‐infrared spectral quantitative analysis

A consensus successive projections algorithm – multiple linear regression method for analyzing near infrared spectra

Contact Info

Product

Resources

About