Performance Comparative Analysis of Decision Tree and Logistic Regression Algorithms for Dry Beans Prediction
Abstract
Dry beans are the most generally developed palatable vegetable harvest around the world, with high hereditary variety. Crop creation is firmly affected by seed quality. Thus, seed characterization is significant for both showcasing and creation since it helps fabricate economical cultivating frameworks. The accurate prediction of dry beans can greatly benefit agricultural practices by enabling effective crop management. In this study, we compared the performance of two popular machine learning algorithms, Decision Tree and Logistic Regression, for dry beans prediction. Experimental results showed that the Decision Tree algorithm achieved an accuracy of 95.81%, with precision and recall values of 95.9% and 95.8% respectively. On the other hand, Logistic Regression achieved an accuracy of 92.79%, with precision and recall values of 92.8%. The Decision Tree algorithm outperformed Logistic Regression in all metrics, showcasing its ability to accurately classify dry beans instances. These findings contribute to the growing body of knowledge in machine learning for agricultural applications, providing valuable insights for optimizing crop management strategies.
Keywords
Download Options
Introduction
Individuals eat dry beans, which are a kind of vegetable that is self-pollinated. Beans are a critical harvest on a worldwide scale and are well known with the two ranchers and customers. Dry beans represent almost 50% of the grain vegetables consumed straight by people in most of agricultural nations [1][2]. An arrangement of value control ensures that endorsed seed meets public and worldwide quality benchmarks. For most of food items, visual attributes are the essential measure utilized by customers while going with buying choices [4]. Like other vegetable species, normal beans show the most variety as far as development designs, actual elements (size, shape, and concealing), development, and capacity to develop and adjust [11]. Arranging and ordering bean seeds physically is a tedious interaction. Moreover, this technique is wasteful and dreary, especially while working with enormous creation volumes. Human examiners are generally accountable for actually looking at unrefined components, and smoothing out the auditors' findings is troublesome. These contemplations reaffirm the significance of true estimation frameworks. Thus, programmed evaluating and order strategies are required.
Conclusion
In conclusion, the experimental results demonstrate the effectiveness of both Decision Tree and Logistic Regression algorithms for dry beans prediction. The Decision Tree algorithm exhibited superior performance in terms of accuracy, precision, and recall, making it a strong candidate for further research and implementation in real-world scenarios. These findings contribute to the growing body of knowledge in machine learning for agricultural applications, potentially aiding in crop management and optimizing agricultural practices