RESUMO
Spectral preprocessing techniques can, to a certain extent, eliminate irrelevant information, such as current noise and stray light from spectral data, thereby enhancing the performance of prediction models. However, current preprocessing techniques mostly attempt to find the best single preprocessing method or their combination, overlooking the complementary information among different preprocessing methods. These preprocessing techniques fail to maximize the utilization of useful information in spectral data and restrict the performance of prediction models. This study proposed a spectral ensemble preprocessing method based on the rapidly developing ensemble learning methods in recent years and the ridge regression (RR) model, named stacking preprocessing ridge regression (SPRR), to address the aforementioned issues. Different from conventional ensemble learning methods, the proposed SPRR method applied multiple different preprocessing techniques to the original spectral data, generating multiple preprocessed datasets. These datasets were then individually inputted into RR base models for training. Ultimately, RR still served as the meta-model, integrating the output results of each RR base model through stacking. This approach not only produced diversity in base models but also achieved higher accuracy and lower computational complexity by using a single type of base model. On the apple spectral dataset collected by our team, correlation analysis showed significant complementary information among the data produced by different preprocessing techniques. This provided robust theoretical support for the proposed SPRR method. By introducing the currently popular averaging ensemble preprocessing method in a comparative experiment, the results of applying the proposed SPRR method to six datasets (apple, meat, wheat, olive oil, tablet, and corn) demonstrated that compared to the single preprocessing method and averaging ensemble preprocessing method, SPRR yielded the best accuracy and reliability for all six datasets. Furthermore, under the same conditions of the training and test datasets, the proposed SPRR method demonstrated better performance than the four commonly used ensemble preprocessing methods.
RESUMO
Inspired by aquaphotomics, the optical path length of measurement was regarded as a perturbation factor. Near-infrared (NIR) spectroscopy with multi-measurement modals was applied to the discriminant analysis of three categories of drinking water. Moving window-k nearest neighbor (MW-kNN) and Norris derivative filter were used for modeling and optimization. Drawing on the idea of game theory, the strategy for two-category priority compensation and three-model voting with multi-modal fusion was proposed. Moving window correlation coefficient (MWCC), inter-category and intra-category MWCC spectra, and k-shortest distances plotting with MW-kNN were proposed to evaluate weak differences between two spectral populations. For three measurement modals (1 mm, 4 mm, and 10 mm), the optimal MW-kNN models, and two-category priority compensation models were determined. The joint models for three compensation models' voting were established. Comprehensive discrimination effects of joint models were better than their sub-models; multi-modal fusion was better than single-modal fusion. The best joint model was the dual-modal fusion of compensation models of one- and two-category priority (1 mm), one- and three-category priority (10 mm), and two- and three-category priority (1 mm), validation's total recognition accuracy rate reached 95.5%. It fused long-wave models (1 mm, containing 1450 nm) and short-wave models (10 mm, containing 974 nm). The results showed that compensation models' voting and multi-modal fusion can effectively improve the performance of NIR spectral pattern recognition.
