Quality control of fresh strawberries by a random forest model

Abstract

BACKGROUND Strawberry quality is one of the most important factors that guarantees consistent commercialization of the fruit and ensures the consumer's satisfaction. This work makes innovative use of random forest (RF) to predict sensory measures of strawberries using physical and physical‐chemical variables. Furthermore, it also employs these same physical and physical‐chemical variables to classify strawberries in the classes "satisfied" or "not satisfied" and "would pay more" or "wouldn't pay more. The RF‐based model predicts the acceptance, expectation, ideal of sweetness, ideal of acidity, and the ideal of succulence based on the physical and physical‐chemical data. Then, the predicted parameters are used as input for the RF‐based classification model. RESULTS The RF achieved a coefficient of determination R2 > 0.72 and a root‐mean‐squared error (RMSE) smaller than 0.17 for the prediction task, which indicates that one can estimate the sensory measures of strawberries using physical and physical–chemical data. Furthermore, the RF was able to classify 87.95% of the strawberry samples correctly into the classes ‘satisfied’ and ‘not satisfied’ and 78.99% in the classes ‘would pay more’ or ‘would not pay more’. A two‐step RF model, which employed both physical and physical–chemical data to classify strawberry samples regarding the consumer's response also correctly classified 100% and 90.32% of the samples with respect to consumers’ satisfaction and their willingness to pay more, respectively. CONCLUSION The results indicate that the developed models can be used in the quality control of strawberries, supporting the establishment of quality standards that consider the consumer's response. The proposed methodology can be extended to control the sensory quality of other fruits. © 2021 Society of Chemical Industry

A comparison of machine learning algorithms for predicting consumer responses based on physical, chemical, and physical–chemical data of fruits

“…Sample preparation and pre‐processing followed the methodology employed by Corrêa et al (2014) and Ribeiro et al (2021). Initially, the fruits were hand‐picked to discard those which had visible physical damage.…”

“…The physical and physical–chemical analysis were performed in six repetitions for every fruit sample, strictly following the methodology presented by Ribeiro et al (2021). The titratable acidity (TA), soluble solids (SSC), and the SSC/TA ratio were computed following the Latimer (2019) guidelines.…”

“…Furthermore, each session was performed using a balanced complete block design. The sensory analysis strictly followed the methodology employed by Ribeiro et al (2021). We evaluated the acceptance, expectation, ideal of sweetness, ideal of succulence, and the ideal of acidity of each fruit.…”

“…However, these attributes, if unaccompanied, are not good indicators of the fruits’ quality (Corrêa et al, 2014). Therefore, this visual‐based selection can potentially lead to consumer's dissatisfaction as the chosen fruit may not be as good as it looks (Ribeiro et al, 2021; Yu, Low, & Zhou, 2018).…”

“…Using such algorithms, it is possible to predict consumer responses accurately. Thus, consumers can be aware of the fruit quality beforehand, which narrows their expectations and probably diminishes their dissatisfaction (Corrêa et al, 2014; Ribeiro et al, 2021; Yu et al, 2018).…”