Predicting Durian Pulp Weight and Loss Using Classification Methodologies: A Comparative Study

Agricultural technology plays a crucial role in promoting sustainable growth within the food industry and the overall business cycle. This research focuses on developing a method to predict the pulp weight and weight loss of durians, to advance agricultural technology. We evaluated the effectiveness of four classification methodologies: Decision Tree, Naive Bayes, Deep Learning, and Random Forest, by comparing their accuracy in predicting these characteristics based on durian property data. Our findings indicate that these methodologies a re more successful in predicting pulp weight than weight loss. Furthermore, combining multiple methodologies generally resulted in higher accuracy compared to using a single methodology. This study contributes to the development of future agricultural technologies by providing insights into the potential of various classification methods for predicting important d urian qualities. These predictions can help farmers and retailers make informed decisions, ultimately leading to a more efficient a nd sustainable agricultural sector.