摘要
为使城市环境空气质量评价方法具有更高的准确性和鲁棒性,提出将随机森林模型引入城市环境空气质量评价中的方法。通过对随机森林模型进行训练,建立起空气质量评价因子与空气质量等级之间的内在映射关系。基于随机森林模型的评价方法,对上海市空气质量数据进行评价实验,实验结果表明,评价预测平均准确性稳定在99.06%,当树数目为16个时,效果最好,准确性可达99.69%。实例验证了该方法的可行性。
To improve the accuracy and robustness of evaluation of ambient air quality,a method of introducing the random forest model into the air quality evaluation was presented.By training the random forests model,the internal mapping relationship between the grades of air quality and the air quality index was established.Experiments were conducted in the data set of air quality of Shanghai based on random forests model.Experimental results show that the evaluation and prediction accuracy stabilize at an average value of 99.06% using the proposed method.When the number of trees is 16,the accuracy reaches the optimal result of 99.69%.The feasibility of the method is verified by the application.
作者
杨瑞君
赵楠
凡耀峰
侯梅芳
YANG Rui-jun ZHAO Nan FAN Yao-feng HOU Mei-fang(School of Computer Science and Information Engineering, Shanghai Institute of Technology, Shanghai 201418,China Ecological Technique and Engineering College, Shanghai Institute of Technology, Shanghai 201418,China)
出处
《计算机工程与设计》
北大核心
2017年第11期3151-3156,共6页
Computer Engineering and Design
基金
国家自然科学基金项目(41571044)
中国气象局气候变化专项基金项目(CCSF201503)
关键词
随机森林模型
空气质量评价
分类
泛化误差
空气质量指数
random forest model
urban air quality evaluation
classification
generalization error
air quality index
