一种基于决策树算法的耕地地力等级评价

摘要

An evaluation model of soil fertility grade for cultivated land was established by decision tree algorithm using the patch data from Dehui City, Jilin Province as experimental data and organic matter, total nitrogen, available phosphorus and available potassium as evaluation indices. The grading result was basically consistent with the actual status, indicating that the method is efficient and feasible for evaluation of soil fertility grade for cultivated land.%以吉林省德惠市的图斑数据作为实验数据,选择有机质、全氮、速效磷、速效钾4个方面作为耕地地力等级的评价指标,利用数据挖掘中的决策树分类技术建立耕地地力等级评价模型.通过实验分析,得到的分级结果与实际情况基本一致,表明该方法对耕地地力等级评价具有有效性和可行性.