Authentication of the Geographical Origin of Shandong Scallop Chlamys farreri Using Mineral Elements Combined with Multivariate Data Analysis and Machine Learning Algorithm

Explainable machine learning for predicting the geographical origin of Chinese Oysters via mineral elements analysis

“…The results of the One-Way ANOVA test indicate significant differences in the concentrations of 18 elements (Na, Mg, Ca, K, Al, Fe, Zn, Sr, Mn, V, As, Ag, Se, Cd, Li, Ba, Co, and Bi). Similarly, significant differences in the profiling of Al, Mn, Fe, Cu, Zn, V, Cr, Co, Ni, As, Sr, Cd and Pb were detected in a research of scallop origin traceability ( Kang et al, 2022 ). The average concentrations of As, Li, V, Sr, K, Mg, and Na in oysters from the northern regions of China (specifically Liaoning and Shandong Provinces) were significantly higher ( p < 0.05) than those in oysters from the southern regions of China (specifically Guangdong and Guangxi Provinces).…”

“…To mitigate adverse effects and prevent data leakage, we selected the Standard Scaler , applied the fit() function to the training set, and subsequently used the transform() function on both the training and testing sets to normalize each dataset. Moreover, the synthetic minority oversampling technique (SMOTE) was exclusively applied to the training set ( Kang et al, 2022 ).…”

“…Among these technologies, elemental fingerprinting stands out as a fast, cost-effective, and straightforward method that relies on the understanding that the elemental composition of the maricultural environment can be accurately represented in the organisms themselves ( Kang et al, 2021 ). Elemental fingerprinting has proven successful in determining the origin of Asian seabass, Black tiger prawns, sea cucumber and scallops ( Honig et al, 2020 ; Han et al, 2022a ; Kang et al, 2022 ). However, the suitability of this method for tracing the origin of Chinese oysters is questionable.…”

“…Machine learning has emerged as a valuable tool for investigating food security, food fraud and food origin traceability research, even when traditional face-to-face survey data is lacking ( Jiménez-Carvelo et al, 2019 ; Deng et al, 2021 ; Kang et al, 2022 ; Martini et al, 2022 ; Huang et al, 2023 ). Machine learning has proved to be effective in various fields, including economic, biology, manufacturing, and health promotion research, primarily due to its strong classification performance after alleviating the influence of overfitting by data reduction, feature selection, cross-validation and other measures ( Lundberg and Lee, 2017 ; Lundberg et al, 2020 ; Kim and Kim, 2022 ).…”

Rastreabilidade da carne bovina brasileira avaliada por técnicas isotópicas e quimiométricas

“…Avanços no controle da qualidade de alimentos utilizando técnicas quimiométricas têm sido foco de estudo de diversos autores nos últimos anos [108][109][110][111][112]. Algoritmos de classificação que implementam a aprendizagem supervisionada, como Random Forest, Sequential Minimal Optimization, Naïve Bayes, Multilayer Perceptron e Classification and Regression Tree, foram utilizados em estudos de autenticação de café orgânico [113], origem geográfica do café [114], diferenciação de cana-de-açúcar e sucos de produção orgânica ou convencional [115,116],…”