Modeling quality changes in Pacific white shrimp (Litopenaeus vannamei) during storage: Comparison of the Arrhenius model and Random Forest model. Issue 1 (18th November 2020)

• Record Type: Journal Article
• Authors: Chen, Shengjun; Tao, Feiyan; Pan, Chuang; Hu, Xiao; Ma, Haixia; Li, Chunsheng; Zhao, Yongqiang; Wang, Yueqi
• Is Part Of: Journal of food processing and preservation. Volume 45:Issue 1(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-11-18
• Subjects: Arrhenius model -- deterioration -- Pacific white shrimp -- Random Forest model; Food -- Preservation -- Periodicals; Food industry and trade -- Periodicals; 664.005
• DOI: 10.1111/jfpp.14999
• Languages: English
• ISSNs: 0145-8892

Abstract

Abstract: To investigate the quality changes of Pacific white shrimp stored at 4, −3, and −20°C, indicators including sensory assessment, pH, texture (hardness, springiness, gumminess, chewiness), thiobarbituric acid (TBA), total sulfhydryl content, Ca 2+ ‐ATPase activity, and total viable counts (TVC) were studied in this work. The Random Forest model was chosen to estimate the quality changes of these indicators in comparison with the Arrhenius model. During different temperatures, pH, TBA, and TVC increased with the extension of storage time, while the other indicators decreased. Compared with the Arrhenius model, the relative errors of quality indicators of the Random Forest model were below 10%, r 2 was close to 1, and root mean square error was mostly below 0.1, which meant a better fitting property for these indicators. Thus, the Random Forest model with higher prediction accuracy is a hopeful method for predicting the changes in the quality of Pacific white shrimp. Practical applications: The Random Forest model provides a more accurate and convenient model to predict the quality changes of Pacific white shrimp under the temperature range from −20°C to 4°C, which shows a potential use for shrimp preservation and processing. Random forest model cannot only be used for estimating soil calcium carbonate and other regression issues, but also assessing the shelf life by predicting the values of quality indicators of aquatic products during its storage.