摘要： 提出一种以J48决策树算法为基础的水质评价方法。通过分析K最近邻法、贝叶斯分类法、向量空间模型法等常用的水质评价方法，针对某地区2008年的水质监测数据，采用J48决策树算法和单因子污染指数法对该地区水质进行计算，并对水质分类数据进行分类、属性清理和J48决策树参数的选择，以此建立模型。实验结果表明，该方法的交叉验证率约为95%，具有较好的分类评估效果。

关键词: 水质评价, 属性清理, 交叉验证, 单因子污染指数法, 决策树

Abstract: This paper proposes a water quality evaluation method based on J48 decision tree algorithm. By analyzing the common methods of water quality evaluation such as K nearest neighbor, Bayesian classification, vector space model, based on the water quality monitoring data of the region in 2008, it calculates the water quality of one region by using J48 decision tree algorithm and single-factor pollution index. A model is set up after classifying dates, cleaning up attributes and choosing parameters of J48 decision tree. Experimental result shows that the validation rate of J48 decision tree algorithm is about 95%. It shows that this method which this paper chooses has a good classifying assessment effect.

Key words: water quality evaluation, attribution cleaning, cross-validation, single factor pollution index method, decision tree

中图分类号: 

• TP301.6