基于改进的C4.5算法对玉米病虫害治理方案分类研究

Research on classification of maize pests and diseases management programs based on improved C4.5 algorithm

通过机器学习,采用特定的算法模型对农作物病虫害治理方案进行匹配,这是现代农业的发展方向之一。针对C4.5算法匹配方案时存在的精确度低、大容量数据集处理效率差等问题,提出一种改进的C4.5算法。改建的算法采用K-means++算法对数据进行离散化处理,根据特征标签值概率对测试集缺失值数据按权生成,并基于梯度优化训练集与聚类中心点。实验结果表明,改进的C4.5算法面对不同容量测试集,能够在保证效果良好的前提下,提高决策精度与治理效率。

Matching maize pest control programs with specific algorithmic models by machine learning is one of the development directions of modern agriculture.Aiming at the problems of low accuracy and poor processing efficiency of large-capacity datasets when matching programs with C4.5 algorithm,an improved C4.5 algorithm is proposed.It uses the K-means++algorithm to discretize the data,generates the missing value data of the test set by weight according to the probability of the feature label value,and optimizes the training set and clustering center point based on the gradient.The experimental results show that the improved C4.5 algorithm effectively improves the decision-making accuracy and management efficiency for different capacity test sets,while ensuring good performance.