基于XGBoost的储粮品质预测  

Prediction of stored grain quality based on XGBoost

作　　者：张慧媛 曾显超 钟克针 韩帅 唐怀建[1] Zhang Huiyuan;Zeng Xianchao;Zhong Kezhen;Han Shuai;Tang Huaijian(School of Food and Strategic Reserves,Henan University of Technology,Zhengzhou,Henan 450001;School of Artificial Intelligence and Big Data,Henan University of Technology,Zhengzhou,Henan 450001)

机构地区：[1]河南工业大学粮食和物资储备学院,河南郑州450001 [2]河南工业大学人工智能与大数据学院,河南郑州450001

出　　处：《粮食科技与经济》2023年第6期67-71,98,共6页Food Science And Technology And Economy

摘　　要：研究了基于机器学习的粮食品质预测技术,创新性地采用机器学习中的XGBoost算法进行粮情检测,通过对粮仓中温度、湿度和化学物质含量进行特征向量分析,旨在帮助粮仓管理人员更好地了解储存粮食的状态,预测可能的变质和虫害情况,保证粮食质量和安全。针对算法检测的准确率,通过XGBoost和传统的机器学习分类算法支持向量机和逻辑回归算法对实验粮食数据进行预测并对比,结果表明,机器学习XGBoost、支持向量机、逻辑回归等算法在粮情检测上都可以快速且正确地做出预测,XGBoost算法的结果相对于支持向量机和逻辑回归,具有更高的准确率。XGBoost算法在提高粮食品质监测效率和准确性方面具有较大潜力。The grain quality prediction technology based on machine learning was studied,the XGBoost algorithm in machine learning was innovatively used to detect grain conditions,and the quality of grain was judged by analyzed the eigenvectors of temperature,humidity and chemical substance content in the granary.It was designed to help granary managers to better understand the status of stored grain,predict possible deterioration and pest damage,and ensure grain quality and safety.Aiming at the accuracy rate of algorithm detection,XGBoost and traditional machine learning classification algorithm supporting vector machine and logistic regression algorithm were used to predict and compare the experimental grain data.The empirical results showed that machine learning algorithms such as XGBoost,support vector machine and logistic regression can quickly and correctly make predictions in food situation detection.Compared with support vector machines and logistic regression,the XGBoost algorithm had higher accuracy in grain situation detection.The XGBoost method have great potential in improving the efficiency and accuracy of grain quality monitoring.

关 键 词：储粮品质 逻辑回归 支持向量机 XGBoost算法 预测模型 

分 类 号：TS210.2[轻工技术与工程—粮食、油脂及植物蛋白工程]