Research on Grading Method of Korla Fragrant Pears Based on Gradient Boosting Decision Tree

Read full text Summarise with VIRAT
Authors: Shuting Xu, Shuo Yang, Mengyuan Tao, Yixin Gao, Zhenghao Cui, Yingchao Wang
Cited by: 0
References: 9

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

To address the problems of traditional Korla fragrant pear grading relying on manual experience, strong subjectivity, low efficiency and insufficient grading accuracy, an intelligent grading method for Korla fragrant pears based on Gradient Boosting Decision Tree (GBDT) was proposed. Taking Korla fragrant pears as the research object, the grading indicators were determined in accordance with Xinjiang local standards and group standards, and four core features including single fruit weight, peeled fruit firmness, soluble solid content and fruit surface defect degree were selected as the model input to construct a GBDT model for pear grading. A total of 360 pear sample data were collected on site, and after data preprocessing and feature optimization, the model was trained, verified and parameter-tuned. Comparative experiments were conducted with traditional machine learning algorithms such as Random Forest (RF) and Support Vector Machine (SVM). The results showed that the grading accuracy of the constructed GBDT grading model on the test set reached 95.83%, which was higher than that of the RF model (91.67%) and the SVM model (88.33%). Moreover, it performed better in terms of recall rate and precision rate for the grading of super grade, first grade and second grade fragrant pears, and could effectively distinguish Korla fragrant pears of different grades. This method realizes the objective, accurate and efficient grading of Korla fragrant pears, and provides technical support and theoretical reference for the intelligent sorting of the fragrant pear industry.

Fields of Study & Subjects

Smart Agriculture and AI Agricultural Engineering and Mechanization Data Mining and Machine Learning Applications

Publication details

Journal / source
Academic Journal of Science and Technology
Volume
20
Issue
1
Pages
140–143
Publication date
26 March 2026
Language
EN