Machine learning algorithms to estimate drying characteristics of apples slices dried with different methods

SAĞLAM C., ÇETİN N.

Journal of Food Processing and Preservation, cilt.46, sa.10, 2022 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 46 Sayı: 10
  • Basım Tarihi: 2022
  • Doi Numarası: 10.1111/jfpp.16496
  • Dergi Adı: Journal of Food Processing and Preservation
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Aquatic Science & Fisheries Abstracts (ASFA), Biotechnology Research Abstracts, Business Source Elite, Business Source Premier, CAB Abstracts, Compendex, Food Science & Technology Abstracts, INSPEC, Veterinary Science Database
  • Kayseri Üniversitesi Adresli: Hayır

Özet

© 2022 Wiley Periodicals LLC.In this study, three different apple cultivars were dried using five different drying methods and moisture ratio (MR), moisture content (MC) and drying rate values were determined. Then, different machine learning algorithms (artificial neural network, k-nearest neighbors, random forest, gaussian processes, and support vector regression) were used to estimate MR, MC and drying rate. For MR estimation of Golden Delicious, Oregon Spur and Granny Smith cultivars, Random Forest was most successful algorithm with correlation coefficients (R) of 0.9800, 0.9873, and 0.9841, respectively. This was followed by SVR with R: 0.9323 for Golden Delicious, ANN with R: 0.9766 for Oregon Spur and 5-NN with R: 0.9827 for Granny Smith. MC and drying rate estimation results showed that RF, SVR, and k-NN achieved higher R for all cultivars. It was concluded that machine learning algorithms are an effective approach for the accurate estimation of the drying characteristics of apple slices. Practical applications: Machine learning-like precise modeling techniques are used to estimate the drying characteristics of agricultural commodities. Models should be assessed and compared for optimization of drying conditions and operational costs. Machine learning models predictions agreed well with testing data sets and they could be useful for understanding and controlling the factors affecting drying behaviors.