We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Physicochemical characterization of Spanish cherry (Mimusops elengi) fruit at different growth stages and its mass modelling using machine learning algorithms.
- Authors
Srivastava, Prashant Kumar; Sit, Nandan
- Abstract
Spanish cherry (Mimusops elengi) is an underutilized fruit with extensive nutritional and therapeutic properties. The quality of fruit is directly linked to its physical and biochemical characteristics, which are crucial for its utilization and consumer acceptance. This study aims to investigate the physical, biochemical, thermal, colour, and functional properties of Spanish cherry fruits. Additionally, it aims to predict the mass of the fruit at different growth stages (young, premature, mature, pre-ripe, ripe) using various machine learning algorithms, including multilayer perceptron, linear regression, support vector regression, and Gaussian process. Considering these features will be advantageous in designing and developing equipment for various tasks, including grading, sizing, peeling, storage, packing systems, and preservation. All five developmental stages of Spanish cherry fruits exhibited significant (P < 0.05) and sequential changes in their physiological and biochemical properties. The total phenolic content, total flavonoid content, and antioxidant activity of fruits at the young to ripen stages ranged from 304 to 123 mg GAE/100 g, 148 to 88 mg QE/100 g, and 90.8 to 84.3%, respectively. Among the machine learning algorithms, the multilayer perceptron and support vector regression models demonstrated the best fit, with the highest correlation values for predicting the mass of fruits at their respective growth stages. The multilayer perceptron model outperformed other machine learning algorithms, exaggerate the highest correlation coefficient and the lowest RMSE and RRSE values of 0.99, 0.23, and 13.53%, respectively, for predicting the mass of fruit at any developmental stage. This study will be invaluable in providing a unified approach for scientists and local farmers to develop processing machinery that fully harnesses the potential of this fruit crop.
- Subjects
MACHINE learning; FRUIT; FRUIT ripening; CHERRIES; GAUSSIAN processes; SWEET cherry; FLAVONOIDS
- Publication
Journal of Food Measurement & Characterization, 2024, Vol 18, Issue 5, p3906
- ISSN
2193-4126
- Publication type
Article
- DOI
10.1007/s11694-024-02464-3