【摘要】 针对农作物病虫害诊断过程中缺失值会降低分类精度的问题,分析数据集缺失值的填充问题,提出结合贝叶斯理论的决策树分类算法。在分类模型基础上,通过贝叶斯方法预测并填充缺失值对C4.5决策树算法进行改进,用C4.5决策树算法对病虫害分类诊断。将此方法和C4.5算法在不同数据集上进行测试比较,验证了该算法的有效性。更多还原

【Abstract】 For corps pests and diseases,the missing data of corps pests and diseases may reduce the accuracy of the diagnosis and classification.The imputation of missing value of datasets was analyzed and decision tree algorithm based on Bayesian algorithm was studied.Based on the classification model,the C4.5 decision tree algorithm was modified by predicting and imputing the missing value using Bayesian algorithm,and the corps pests and diseases were diagnosed and classified using C4.5 decision tree algorithm.Different datasets were used to verify the presented algorithm and C4.5 decision tree algorithm,demonstrating the effectiveness of the presented algorithm.更多还原