Grading of Mango Fruits Based on Physical Measurements

Harumanis Mango is a popular mango species in Malaysia. In order to handle the quality of the mango, a grading procedure is required. Currently, the grading of the mango is using human labor, which increases the possibility of human error. Therefore, this study proposed a classification of the mango grade by using machine learning with statistical data to reduce the possibility of human error. The machine learning models that have been selected in this study are Support Vector Machines (SVM), Artificial Neural Network (ANN), and K-Nearest Neighbors (k-NN). The result of this study showed that the SVM has higher accuracy, precision, recall, and F1-score value compared to ANN and k-NN. In conclusion, the machine learning model is able to classify the mango grade by using the weight, length, and circumference of the mango in order to help reduce the possibility of human error in the mango grading procedure.