An empowered adaptive neuro-fuzzy inference system using self-organizing map clustering to predict mass transfer kinetics in deep-fat frying of ostrich meat plates

Deep-fat frying (DFF) is a cooking process, in which water containing foodstuff is immersed into edible oils or fats at temperatures above the boiling point of water. This process is a fast and easy method to prepare tasty foods; therefore, despite the trend to low-fat foods, deep-fried products enjoy increasing popularity. Moisture content (MC) and fat content (FC) are very important quality indicators for fried foods in terms of health concerns and palatability of the products. This paper presents a new approach based on an adaptive neuro-fuzzy inference system (ANFIS) and self-organizing map (SOM) clustering for more accurate predicting MC and FC during DFF of ostrich meat plates. First the data set of each mass transfer parameter was categorized into two clusters by SOM method, and at the next stage each cluster was fed into an independent ANFIS models with the ability of rule base extraction and data base tuning. To train the ANFIS prediction system, triangular membership function (MF) was chosen. Results showed that the optimized ANFIS model with clustering improved the prediction ability of ANFIS and truly described mass transfer during the DFF (12.46% improvement with R =  0.96 for MC and 5.46% improvement with R = 0.92 for FC). This methodology can also be applied to optimize the operating conditions.

Research highlights
► Moisture content (MC) and fat content (FC) are very important quality indicators for fried foods in terms of health concerns and palatability of the products.
► This paper presents a new approach based on adaptive neuro-fuzzy inference system (ANFIS) and self-organizing map (SOM) clustering for predicting MC and FC during DFF of ostrich meat plates.
► Results showed that the optimized ANFIS model truly described mass transfer during the DFF process (R = 0.96 for moisture content and 0.92 for fat content).