Performance Improvement of NIR Spectral Pattern Recognition from Three Compensation Models’ Voting and Multi-Modal Fusion

Abstract

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.

Aquaphotomics—Exploring Water Molecular Systems in Nature

“…Out of the 21 published articles, 3 were reviews [ 2 , 3 , 4 ] and 18 were original research papers [ 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 ], that can be roughly separated into 6 categories ( Figure 3 ).…”

“…The majority of the articles, approximately one third, belongs to the field of applications in agriculture and food science [ 8 , 9 , 12 , 13 , 14 , 21 , 22 ], a field from which aquaphotomics originated, and probably the reason why this category is still at the forefront of aquaphotomics development. Following this category are the preprocessing and chemometrics [ 2 , 4 , 15 , 18 ] and fundamental studies [ 5 , 6 , 10 , 17 ]. The former category reflects an important issue that aquaphotomics is trying to resolve, namely, the necessity of developing appropriate tools for dealing with the spectra of highly aqueous samples, and that significant efforts are being invested in resolving the issues of extracting the information from very subtle spectral signals.…”

“…Furthermore, two research groups presented the results, in parallel, for both the first overtone of water region and higher overtones, demonstrating increased efforts in gaining more clarity in the nature of information that can be gained from each of the regions, and attempts at defining WAMACs similar to the existing ones in the first overtone [ 23 ]. Additionally, three studies were performed over the entire (~750–2500 nm) [ 16 , 18 ], or nearly the entire range, of the NIR spectrum (1100–2500 nm) [ 15 ]. This testifies to increased interest in the water spectral features over the entire range, which may result in many benefits for both scientific and practical aspects.…”

“…In one of the very innovative and creative works, Ye et al developed a new discriminating analysis for the purpose of application in problems of the classification and identification of water samples with small differences in the component content; this is a very hot topic, especially in the quality control of drinking water [ 18 ]. The work is rich in novelty in many aspects: they utilized the optical path length of cuvette as a perturbation factor, performed fusion modeling on multi-optical path measurements, and developed a new pattern recognition method that effectively discriminated between three types of drinking water with a very similar content.…”